Operational registry

Local skill execution truth board

This is the local Windows Codex shelf only. Each of the 273 active skills has a current bounded receipt. It is not a claim that provider accounts, client sites, or publishing authority are already available.

96PASS

The stated safe local core ran and produced a receipt-backed result.

40PARTIAL

Some local setup works; the receipt names the remaining repair.

137BLOCKED

The local skill is installed, but its real task needs a target, authority, account, provider, or named input.

What this board proves

Local execution status and the exact receipt finding for every skill, without pretending a client or provider task ran when its gate was unavailable.

What it does not prove

Access to WordPress, Google Business Profile, ads, paid APIs, private repositories, or permission to publish or spend.

Why the 137 BLOCKED skills are blocked

BLOCKED does not mean the installed skill is broken. Each card below still carries its exact receipt finding. This is the repair queue grouped by the gate that stopped its current bounded run.

4

Local Windows runtime

Repair the named local runner, browser, package, or Windows permission, then re-run the same bounded task.

69

Provider, credential, or spend

Map the approved environment-variable reference and obtain a bounded provider budget or call approval.

6

Missing source or task input

Supply the exact URL, file, brief, question, competitor, or dated search-source packet the skill names.

38

Target, account, or authority

Name the target and grant the minimum scoped read or write authority required for that one task.

20

No scoped job selected

Choose the task scope and the intended output destination. No account access is needed until the card says so.

Current local execution state

96
PASS
40
PARTIAL
137
BLOCKED

Registry source: canonical local ledger · Receipt index: 273 records · Ledger SHA-256: 2af76e656c73bcb04e165598778ec6b27cf1be34576e2ed62fbb682b5fa1d37a

PASS

96 skills

Local core proven

ad-designer

Create and review static paid-social or display ad images from an approved brief, exact copy, and brand assets. Use for Meta feed, Stories, Reels covers, carousel hooks, price-anchor ads, and other static campaign creative. Do not use for video, logos, or general editorial imagery.

PASSWindows: WINDOWS READY
spielbergskillagent task

Receipt findingThe static-asset core completed on a declared non-client fixture: a text-free background, deterministic exact-copy overlay, 1080 by 1080 PNG, and full-size visual inspection. No campaign was published.

ai-content-detector

Use this skill when analyzing text to determine if it was written by AI or a human. Runs a 5-layer analysis covering word frequency, sentence structure, coherence, emotional depth, and provides a final AI vs human percentage score. TRIGGER when: 'AI detection', 'is this AI written', 'AI sniffer', 'check if AI', 'human or AI'.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingThe local five-layer heuristic report completed on a declared 108-word fictional sample with cited signals and an explicit low-confidence caveat. It is not evidence about any real author's identity.

ai-llm-seo

Use when 'AI SEO', 'LLM optimization', 'ChatGPT citations', 'AI Overviews', 'Perplexity SEO'. Optimizes content for AI search engine visibility.

PASSWindows: WINDOWS READY
einsteinwrapperagent task

Receipt findingThe Content Brief core completed from a fully declared fictional fixture with a 45-word declarative capsule plus entity, schema, distribution, and monitoring plan. It expressly cannot be used publicly until fixture values are replaced with authorized evidence.

analytics-tracking

Use when the user explicitly asks about analytics implementation, trigger on phrases like "set up tracking," "GA4," "Google Analytics," "GA4 property setup," "conversion tracking," "event tracking," "form submission tracking," "phone call tracking," "scroll depth tracking," "UTM parameters," "tag manager," "GTM," "dataLayer," "GTM triggers," "Enhanced Measurement," "custom dimensions," "key events," "audiences," "Measurement Protocol," "analytics implementation," "tracking plan," "DebugView," or wiring CallRail/CallTrackingMetrics/WhatConverts call data into GA4. For A/B test measurement, see /ab-test-setup.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe tracking-plan core completed on a controlled fictional fixture with four events, three dimensions, two proposed conversions, and explicit PII, consent, and validation boundaries. No implementation occurred.

article-rewriter

Rewrite/uniquify an existing article for SEO. Triggers: rewrite article, spin this, reword, unique content, SEO rewrite, pass copyscape.

PASSWindows: WINDOWS READY
einsteinskillagent task

Receipt findingThe documented rewrite task completed on a generic fixture, producing a 3,280-word Markdown article that passed the installed skill's explicit length and hierarchy requirements.

auto-schema-from-template

Generate valid JSON-LD from a webpage or an existing template using verified business, service-area, entity, and service data. Use for schema markup, JSON-LD, structured data, and schema-from-template requests.

PASSWindows: WINDOWS READY
ghostskillagent task

Receipt findingThe core local template generation and policy checks produced valid fictional JSON-LD with meters-based service radius. It is deliberately not presented as deployable client markup.

baoyu-image-gen

Generate or edit images with the bundled TypeScript pipeline and any implemented provider for which credentials resolve. Use for text-to-image, reference-image edits, aspect-ratio control, and approved batch image generation.

PASSWindows: WINDOWS READY
picassoskillrunner

Receipt findingThe documented Codex-native image route produced one inspected 1672 by 941 generic fixture PNG with no text, logo, people, watermark, client identity, or commercial claim.

before-after-image-generator

Use when generating matching before/after image PAIRS where only cleanliness changes (identical room, furniture, angle) for local service businesses (cleaning, restoration, junk removal, maid service, remodel). Triggers: 'before after images', 'before/after slider images', 'before after pair', 'dirty to clean room', 'messy to spotless', 'before after for [room]', 'matching before/after'.

PASSWindows: WINDOWS READY
picassowrapperagent task

Receipt findingGenerated a matched clean image, image-edited it into a dirty version, then ran the shipped local compositor. The unpublished labelled pair passed visual inspection.

bio-description

Extracts NAP, writes keyword-rich company bio from a URL. Triggers: bio description, company description, extract business info, NAP from URL.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe local identity-extraction and description task completed from a declared offline fixture, producing a 294-character description and five-line address. It is not a live URL scan or real business record.

brand-assets-setup

Use when setting up a project's brand infrastructure folder: 'set up brand assets', 'create brand folder', 'brand assets setup', 'set up project branding', 'design tokens folder', 'create brand guide files', or starting a new web project after brand-dna/onboarding. Creates brand_assets/ with brand-guide.md, colors.css tokens, logos/, fonts/, BRAND_README.md for on-brand decisions. NOT for brand identity strategy from scratch (brand-guideline-generator) or brand discovery (brand-dna).

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe isolated local scaffold-creation core completed inside its evidence folder only. It does not create a usable client brand package.

brian-winum-llms-amplifier

WP plugin, llms.txt/llms-full.txt, OKF bundles. Triggers: LLMS Amplifier, llms.txt, Brian Winum.

PASSWindows: DISPATCH INPUT REQUIRED
schemabrainskillagent task

Receipt findingThe local staged-artifact availability core verified the plugin entry point, content guide, and header example. No file modification, WordPress, client, provider, live endpoint, publishing, or spend action occurred.

browser-lane-router

Choose the safest available browser lane for authenticated or interactive web work.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe local routing core selected the documented direct HTTP lane for a controlled public-robots fixture, stated a one-read-only-GET boundary, and explicitly rejected browser, automation, provider, authentication, proxy, CAPTCHA, and secret access.

chaser-cli

Use when 'Chaser', 'Chaser task', 'send Slack task', 'create task in Slack'. Wraps Chaser REST API to create/list/update/complete Slack tasks from the CLI, no MCP, no OAuth.

PASSWindows: WINDOWS READY
merlinwrapperrunner

Receipt findingThe documented local chaser.py exists and its help exposed eleven command groups. No credential read, provider call, Slack action, task mutation, or client access occurred.

chrome-ripper

Use when Mike asks for Chrome Ripper, Universal Export Downloader, browser-session export scraping, bulk downloading visible dashboard exports, downloading Git provider repository ZIPs from signed-in pages, terminal export downloads, or reusing the Chrome Ripper extension.

PASSWindows: WINDOWS READY
chrome-ripperwrapperagent task

Receipt findingThe canonical local project was found and its npm download command help completed successfully. No browser profile, session, data, URL download, authentication value, edit, or external request was accessed.

content-formatter

Formats raw text into structured docs with headings and hierarchy. Triggers: format this, clean up content, structure this document.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingA fresh local Codex execution read the installed formatting contract and transformed a supplied Flat Fee test-notes fixture into structured Markdown, retaining the given targets, limits, inputs, pass condition, and blockers.

content-gap-analysis

Compare a site's existing content with competitors, keyword or topical-map data, and audience needs to identify and prioritize missing or weak coverage. Use for content-gap audits, topic opportunities, competitor coverage comparisons, and editorial roadmaps.

PASSWindows: DISPATCH INPUT REQUIRED
ghostskillagent task

Receipt findingThe documented content-gap core workflow completed against a fully supplied local fixture. External demand and performance fields remained explicitly unavailable rather than invented.

content-humanizer

Use this skill when rewriting AI-generated content to sound naturally human-written. Applies conversational tone, varied sentence structure, personal touches, idioms, and imperfections while maintaining readability and SEO value. TRIGGER when: 'humanize', 'make it sound human', 'rewrite AI content', 'content humanizer', 'AI proof'.

PASSWindows: WINDOWS READY
gordonskillagent task

Receipt findingThe documented humanization task completed on a generic fixture, preserving its planning topic and structure while meeting the skill's natural-language variation gates.

content-quality-scorer

Multi-dimensional content quality scoring system. Use when evaluating articles before publishing, checking for AI writing patterns, assessing readability/rhythm, or determining if content meets quality thresholds. Scores across 5 dimensions with a pass/fail threshold of 70.

PASSWindows: WINDOWS READY
gordonskillagent task

Receipt findingThe documented scoring task completed on a synthetic fixture, returning a five-dimension 86/100 report with weighted calculation and specific revision priorities.

content-refiner

Merges SEO recommendations into existing page content. Triggers: refine content, incorporate recommendations, rewrite with suggestions.

PASSWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingA fresh local Codex run rewrote the explicit non-client fixture with one H1, three H2 sections, every supplied recommendation, and no invented client, provider, price, performance, or research claim.

content-variation-engine

Create evidence-backed, genuinely location-specific page variations and measure lexical overlap as one editorial QA signal. Use for multi-location service content, service-area pages, or auditing near-duplicate drafts.

PASSWindows: DISPATCH INPUT REQUIRED
merlinwrapperagent task

Receipt findingThe local variation-and-measurement core completed on a declared controlled fixture with three drafts and measured lexical uniqueness. PASS is limited to the artifact-backed offline run.

copy-editing

When the user wants to edit, review, or improve existing marketing copy. Also use when the user mentions 'edit this copy,' 'review my copy,' 'copy feedback,' 'proofread,' 'polish this,' 'make this better,' or 'copy sweep.' This skill provides a systematic approach to editing marketing copy through multiple focused passes.

PASSWindows: WINDOWS READY
picassowrapperagent task

Receipt findingA fresh local Codex execution read the installed skill and SOP, completed all seven copy-editing sweeps, and delivered lever-tagged rewrites while explicitly rejecting unverified claims in the supplied fixture.

copywriting

Use when user explicitly requests conversion-focused marketing copy for web pages—triggered by domain-specific phrases like 'write homepage copy', 'draft a CTA', 'rewrite landing page headline', 'generate pricing page copy', or 'help with feature page messaging'. Fires only on direct copy creation or revision requests targeting homepage, landing, pricing, or feature pages—never for blog posts, social, ads, or internal docs.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fresh local Codex execution read the installed copywriting skill and SOP, then produced a bounded fictional fixture without accessing a client system, provider, browser, or publishing destination.

cro

Analyze and improve conversion performance for marketing pages and forms. Use for CRO, conversion optimization, low-converting pages, and form abandonment; not for product onboarding or popup-only work.

PASSWindows: RUNTIME CONTRACT REQUIRED
shakespeareskillagent task

Receipt findingThe installed CRO advisory core completed on a clearly labelled fictional fixture, producing bounded recommendations, copy alternatives, and unrun experiment hypotheses with required decision fields.

dataforseo-backlinks

DataForSEO Backlinks API. Triggers, backlink profile, anchor text, referring domains, broken link audit, link gap, spam score.

PASSWindows: WINDOWS READY
einsteinskillrunner

Receipt findingA fully fictional offline fixture completed bare-domain spam and anchor-threshold classification. It is explicitly not a live DataForSEO, backlink, provider, client, or web result.

dataforseo-domain-intel

Domain intelligence via DataForSEO, technology stack detection, CMS identification, marketing/analytics tools, server tech, WHOIS data, and competitor domain discovery.

PASSWindows: WINDOWS READY
einsteinskillrunner

Receipt findingA fully fictional offline fixture completed the documented six-phase OSINT-contract validation. It is explicitly not a live DataForSEO, domain, provider, client, or web result.

dataforseo-gmb

GMB/GBP data and reviews via DataForSEO. Triggers: DataForSEO GMB, GBP data, business reviews, business_data endpoint.

PASSWindows: WINDOWS READY
ghostskillrunner

Receipt findingA fully fictional offline fixture completed asynchronous task-state lifecycle validation. It is explicitly not a live DataForSEO, GBP, provider, client, or web result.

dataforseo-rank-tracking

Rank tracking and domain performance via DataForSEO Labs, ranked keywords (Google/Bing), historical visibility trends, domain rank overview, position distribution, traffic estimates, and keyword movement tracking.

PASSWindows: WINDOWS READY
einsteinskillrunner

Receipt findingA fully fictional offline fixture completed documented volume filtering, top-10 classification, and striking-distance logic. It is explicitly not a live rank-tracking, DataForSEO, provider, client, or web result.

dataforseo-serp

Live Google SERP data via DataForSEO. Triggers: SERP results, who ranks for, local pack, PAA questions, AI Overview, SERP features.

PASSWindows: WINDOWS READY
einsteinskillrunner

Receipt findingA fully fictional offline status-20000 fixture completed local SERP parsing and presentation with two organic rows, two PAA rows, and two features. It is not a live DataForSEO or Google result.

design-shotgun

Design shotgun: generate multiple AI design variants, open a comparison board, collect structured feedback, and iterate. Standalone design exploration you can run anytime. Use when: "explore designs", "show me options", "design variants", "visual brainstorm", or "I don't like how this looks". Proactively suggest when the user describes a UI feature but hasn't seen what it could look like. (gstack)

PASSWindows: DISPATCH INPUT REQUIRED
picassowrapperrunner

Receipt findingThe shipped local design-board builder completed offline with three supplied unpublished image variants, the bounded brief, and feedback-export controls. No client system was changed.

diagram-generator

Use when 'generate diagram', 'architecture diagram', 'workflow diagram', 'org chart'. Outputs Excalidraw JSON, SVG, or ASCII diagrams.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe local diagram-generation core completed with an importable Excalidraw-style five-node flow, four valid bound arrows, and deterministic structural validation.

dream

The daily Dream review — read the last 24h of activity across the user's AI tool stack, identify the four highest-impact prescriptions, and write them to ~/.claude-os/dreams/dream-{date}.json.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe Windows Codex deterministic runner ranked four redacted, local receipt-derived signals, wrote a validated Dream JSON report and adjacent state file, and made no network request.

editorial-article-writer

E-E-A-T articles via AIDA/PAS with citations. Triggers: editorial article, expert article, authority article.

PASSWindows: RUNTIME CONTRACT REQUIRED
picassoskillagent task

Receipt findingThe editorial-writing core completed on a bounded synthetic source packet, producing a 1,426-word article and a 152-character meta description without presenting the fixture as real-world authority.

eeat-audit

Audit content and sites for experience, expertise, authoritativeness, trust, and people-first quality signals. Use for E-E-A-T audits, quality-rater reviews, and helpful-content evaluations.

PASSWindows: WINDOWS READY
einsteinskillagent task

Receipt findingThe installed Analyze phase completed on declared fictional raw content, producing an 8 of 40 evidence-led report and direct verification actions. The low score is only a property of the fixture, not a claim about a real site.

email-sequence

Use when 'email sequence', 'drip campaign', 'welcome series', 'nurture emails'. Multi-step email sequences with subject lines, copy, CTAs, and timing.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingA fresh local Codex execution read the installed skill and produced a bounded fictional email sequence. No external or client action occurred.

fact-check

Fact-check text before publishing. Triggers: fact check, verification, YMYL, social post accuracy.

PASSWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe fact-check core completed on a controlled fictional current-hours claim by correctly classifying it UNVERIFIED with safe wording and a source-verification recommendation. It makes no real-world factual assertion.

featured-snippet-optimizer

Rewrites answers as 40-50 word Featured Snippets. Triggers: featured snippet, snippet optimizer, PAA answer.

PASSWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThree local fixture snippets completed the installed direct-answer format gate at 43, 43, and 41 words. This is not a verified Google featured-snippet opportunity.

fix

Diagnose and fix bugs with the smallest verified change.

PASSWindows: WINDOWS READY
gordonskillrunner

Receipt findingAn in-memory reproduce, diagnose, minimal-guard, and verify fixture reproduced an exact null-method error, identified root cause, and verified null to UNKNOWN plus spaced input to DEMO. No project, remote, or client mutation occurred.

fleet-infra

Resolve fleet machines, Tailscale addresses, SSH routes, and availability.

PASSWindows: DISPATCH INPUT REQUIRED
automaskillrunner

Receipt findingThe skill-provided offline mock self-test passed online, offline, and absent parser assertions with no live Tailscale, remote, provider, credential, or client access.

frontend-animation

Use when 'GSAP', 'Framer Motion', 'Lottie', 'Three.js', 'frontend animation'. Animation library selection and implementation with React patterns.

PASSWindows: RUNTIME CONTRACT REQUIRED
merlinskillrunner

Receipt findingA local CSS-only staggered motion fixture with hover feedback, responsive layout, and reduced-motion support passed static validation. PASS is limited to the bounded local implementation.

frontend-design

Build distinctive production-grade interfaces, dashboards, and web components.

PASSWindows: WINDOWS READY
picassoskillrunner

Receipt findingA self-contained local HTML fixture with accessible controls, responsive rules, and local interactions passed static validation. It is not evidence of a deployed or client-facing build.

geogrid-reporting-sop

GeoGrid visibility tracking and reporting SOP using LocalBrandManager. Setup, scheduling, metric interpretation (AGR/ATGR/Sol-V), weekly reporting workflow, and action triggers for local SEO clients.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe controlled local metric-interpretation core completed without LocalBrandManager access, grid run, credits, or client report generation.

ghl-operator

Operate and inventory GoHighLevel across Mike's agency and all connected sub-accounts. Use for GoHighLevel, HighLevel, LeadConnector, GHL contacts, conversations, calendars, opportunities, pipelines, workflows, funnels, forms, surveys, snapshots, emails, social planner, automations, account comparisons, counts, audits, or location operations. Route through named MCP, official API, authenticated private API, or Mike's existing Chrome tab.

PASSWindows: DISPATCH INPUT REQUIRED
gordonskillrunner

Receipt findingA fully fictional local route-selection fixture completed the documented read-only MCP path after the CLI capability boundary. It is not a GHL account, client, workflow, provider, or mutation result.

git-health

Audit Git integrity, branches, refs, reflogs, and orphaned commits.

PASSWindows: RUNTIME CONTRACT REQUIRED
merlinskillagent task

Receipt findingThe full documented local Git verification sequence completed against the skill-evaluator repository: history, status, branches, remotes, reflog, and a remote fetch with prune. The worktree was clean.

gmb-maps-expert

Audit and improve an eligible Google Business Profile and its supporting local-search presence using accurate business data, policy-compliant reputation practices, and measurable website/citation work. Use for map pack, Google Maps SEO, GBP setup, optimization, and local visibility.

PASSWindows: WINDOWS READY
ghostskillagent task

Receipt findingA policy-safe local preflight completed; unknown business fields remain unavailable and no profile action or business claim occurred.

google-ads-keyword-analyzer

Score, classify, group, and prioritize supplied Google Ads Keyword Planner exports, keyword CSVs, or keyword lists for SEO and content planning. Use when metrics are already provided; use dataforseo-keywords to fetch fresh metrics and google-ads-account-structure for campaign builds.

PASSWindows: RUNTIME CONTRACT REQUIRED
gordonwrapperagent task

Receipt findingThe local classification, grouping, and weighted-score mechanics completed on a fictional three-record fixture at 81.25, 60.75, and 39.25. No Google Ads, provider, client, publication, or spend access occurred.

google-ads-negative-keyword-bank

Use this skill when building or maintaining the negative keyword list for a Google Ads (especially local-services) campaign, before launch or during the weekly Search Terms review. Provides the reusable negative keyword category bank and the build-and-maintain process.

PASSWindows: WINDOWS READY
gordonskillagent task

Receipt findingSix reusable negative-keyword categories were extracted locally without an Ads account, Search Terms report, application, or spend.

google-ads-quality-score

Use this skill when diagnosing or improving Google Ads Quality Score, explaining how Quality Score affects CPC and cost per lead, or auditing the three Quality Score components (Expected CTR, Ad Relevance, Landing Page Experience). Provides the Quality Score cost impact table and the component breakdown.

PASSWindows: DISPATCH INPUT REQUIRED
gordonskillagent task

Receipt findingAn account-free local interpretation of the documented Quality Score cost band and three components completed without an Ads account, application, or spend.

granola-expert

Pull Granola meeting notes, transcripts, attendees, folders. Triggers: Granola, meeting notes, transcript, Granola API.

PASSWindows: WINDOWS READY
merlinskillrunner

Receipt findingThe documented local Granola CLI entry point exists and its help completed without authentication, note access, export, provider call, or client data.

gws-project-manager

Use when 'client project status', 'project tracker', 'standup report', 'cross-service update'. Combines Gmail+Calendar+Drive+Sheets for project mgmt.

PASSWindows: WINDOWS READY
avaskillrunner

Receipt findingA controlled offline status-template fixture validated the documented status vocabulary and table-based reporting shape with no external services accessed.

helpful-content-audit

Audit a page or content set for audience usefulness, evidence, first-hand support, completeness, clarity, trust, and search intent. Use for helpful-content reviews, quality diagnostics, refresh decisions, or prioritized editorial remediation.

PASSWindows: WINDOWS READY
carlosskillagent task

Receipt findingThe documented helpful-content audit completed on supplied synthetic copy, preserving uncertainty where page experience, currentness, or safety evidence was unavailable.

impeccable

Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a frontend interface. Covers websites, landing pages, dashboards, product UI, app shells, components, forms, settings, onboarding, and empty states. Handles UX review, visual hierarchy, information architecture, cognitive load, accessibility, performance, responsive behavior, theming, anti-patterns, typography, fonts, spacing, layout, alignment, color, motion, micro-interactions, UX copy, error states, edge cases, i18n, and reusable design systems or tokens. Also use for bland designs that need to become bolder or more delightful, loud designs that should become quieter, live browser iteration on UI elements, or ambitious visual effects that should feel technically extraordinary. Not for backend-only or non-UI tasks.

PASSWindows: DISPATCH INPUT REQUIRED
picassoskillrunner

Receipt findingThe shipped context runner resolved a real local dashboard target and completed successfully without changes.

incident-response

Triage production incidents, assess blast radius, diagnose, and roll back safely.

PASSWindows: WINDOWS READY
knoxskillrunner

Receipt findingA self-contained fictional no-action SEV1 triage completed with required blast-radius fields, fastest-first diagnosis order, and rollback, hotfix, flag, and mitigation hierarchy. No system or provider call and no root-cause claim occurred.

intent-analyzer

Use when 'intent analysis', 'hidden intent', 'search psychology', 'intent behind intent'. Three-layer intent decoder: observable, inferred, and hidden.

PASSWindows: WINDOWS READY
einsteinskillagent task

Receipt findingThe documented quick-analysis core task completed on a local synthetic fixture. Full analysis remains outside this run because it requires authorized, evidence-backed surface-query collection.

internal-linking-optimizer

Analyze and optimize internal-link architecture, anchor text, orphan pages, topic clusters, crawl depth, and navigation. Use for internal linking, link structure, and orphan-page work.

PASSWindows: WINDOWS READY
einsteinwrapperagent task

Receipt findingThe full manual-data workflow completed against a controlled eight-page fixture, identifying two true orphans and producing specific source-target-anchor repairs, cluster map, navigation recommendations, and phased plan.

kml-generator

Use this skill when generating KML service area maps for local SEO with concentric ring placemarks around a business location. 3-stage workflow covering Python script generation, enrichment via kml-enrichment-web.vercel.app, and Google My Maps import.

PASSWindows: RUNTIME CONTRACT REQUIRED
picassoskillrunner

Receipt findingStage-one local KML generation completed on a fictional fixture, emitting ten rings and 1,493 points. Enrichment and My Maps import were intentionally not invoked.

koray-semantic-seo-master

Apply the Koray Tugberk semantic SEO and COREIS framework for source context, topical authority, retrieval cost, content networks, and topical maps. Use for Koray, semantic SEO, COREIS, and full topical-authority planning.

PASSWindows: WINDOWS READY
ghostwrapperagent task

Receipt findingA controlled fictional framework fixture validated central-entity context, intent, core and outer nodes, and COREIS routing without research, client data, demand claims, or publication.

koray-topical-map

Build topical maps, semantic content networks, core and outer sections, and prioritized topic clusters with the Koray Tugberk methodology. Use for topical maps, content networks, topic clusters, and topical-authority architecture.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingA controlled six-node fictional methodology fixture validated core and outer segmentation, internal-link rules, and illustrative scoring without external research, client data, or publication.

landing-page-generator

Use this skill when creating SEO-optimized landing page content for a local business or service page. Analyzes competitors, maps user intent, and outputs full page content with headings hierarchy, CTAs, FAQs, social proof, and SEO metadata. TRIGGER when: 'landing page', 'service page content', 'landing page copy', 'page content for keywords', 'conversion page'.

PASSWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe installed skill's local content-generation core completed on a clearly labelled fictional fixture, validating hierarchy, CTAs, and metadata limits without live research.

landing-page-optimizer

CRO audit/rewrite of landing page copy, layout, trust signals, CTAs. Triggers: landing page optimize, CRO audit, conversion optimization.

PASSWindows: WINDOWS READY
einsteinskillagent task

Receipt findingThe installed skill's local audit-and-rewrite core completed on a clearly labelled fictional fixture. It is not a client-specific CRO validation.

learn

Manage project learnings. Review, search, prune, and export what gstack has learned across sessions. Use when asked to "what have we learned", "show learnings", "prune stale learnings", or "export learnings". Proactively suggest when the user asks about past patterns or wonders "didn't we fix this before?"

PASSWindows: WINDOWS READY
merlinwrapperrunner

Receipt findingA bounded controlled local read-only fixture returned an empty array and created no store file. No remote provider, client data, publication, or state mutation occurred.

link-building-router

Route ethical link-building, backlink analysis, digital PR, outreach, local citation, internal-linking, and brand-owned content-hub requests to active Codex skills. Use for build links, backlink plans, local links, outreach, and link-gap strategy.

PASSWindows: RUNTIME CONTRACT REQUIRED
picassoskillagent task

Receipt findingThe router core correctly selected outreach-prospecting for a declared generic request while preserving six missing evidence and authorization prerequisites. Routing was not treated as permission to prospect or contact anyone.

long-form-video-seo-2026

Optimize long-form YouTube videos for retention, chapters, transcripts, discoverability, and VideoObject schema. Use for video SEO, watch-time improvement, YouTube chapters, and transcript-led repurposing.

PASSWindows: RUNTIME CONTRACT REQUIRED
spielbergskillagent task

Receipt findingA controlled fictional pre-production fixture validated a 15-second hook, three contiguous query-style chapters, and caption requirement. No asset, research, account, upload, or publishing action occurred.

mainwp

Use when 'MainWP API', 'MainWP REST', 'MainWP monitoring', 'MainWP updates'. REST API calls, auth, monitoring, troubleshooting, bulk operations.

PASSWindows: DISPATCH INPUT REQUIRED
williewrapperrunner

Receipt findingA fully fictional local destructive-operation fixture correctly stayed blocked by safe mode and missing confirmation. It is not a MainWP, site, client, plugin, or mutation result.

maps-lead-scraper

Scrapes Maps pack leads by niche/city, enriches owner email/phone/socials. Triggers: maps leads, lead scraper, GMB leads.

PASSWindows: WINDOWS READY
ghostwrapperrunner

Receipt findingA fully fictional one-row fixture validated all 19 documented lead-sheet columns and honest blank handling for unavailable enrichment. It is not a Maps scrape, enrichment, lead list, or contact-data result.

meta-title-description

3 CTR meta title/description variations, character limits. Triggers: meta title, meta description, title tag, SERP snippet.

PASSWindows: WINDOWS READY
merlinskillagent task

Receipt findingA fresh local Codex invocation read the skill and generated three Flat Fee title/description pairs. Independent validation confirmed title lengths 58, 55, 58 and description lengths 149, 147, 146.

orca-cli

Use the public `orca` CLI to operate Orca-managed worktrees, folder contexts, terminals, repos, automations, worktree comments, and the browser embedded inside the Orca app. Use when the user says "$orca-cli", "use orca cli", "Orca worktree", "child worktree", "cardStatus", "spawn codex/claude in a worktree", "read/wait/send Orca terminal", "terminal send", "full handoff", "handover", "give this to another agent", "another worktree", "Orca browser", or "control the browser inside Orca". Prefer this over raw `git worktree`, ad hoc PTYs, Playwright, or Computer Use when the task touches Orca-managed state. Use Computer Use for browser windows, webviews, or desktop UI outside Orca's embedded browser.

PASSWindows: WINDOWS READY
carloswrapperrunner

Receipt findingThe local Orca CLI returned its version-matched guide and reported the Orca runtime reachable and ready. No worktree, terminal, browser, automation, or mutation was created.

orca-emulator-android

Control an Android emulator / device from inside Orca using the `orca` CLI. Use for listing/booting AVDs, taps, swipes, typing, hardware buttons (incl. Back and Recents), rotation, app install/launch, runtime permissions, the accessibility tree, and logcat — driving a real adb-connected device or emulator. Cross-platform (Windows, Linux, macOS). Complements the orca-emulator (iOS) and orca-cli skills.

PASSWindows: WINDOWS READY
carloswrapperrunner

Receipt findingA read-only local Orca discovery run found five configured Android AVDs, all shut down. No device was created, attached, controlled, or externally accessed.

page-optimizer

On-page SEO audit and rewrite pipeline for an existing URL. Triggers: optimize my page, SEO audit, page optimizer, on-page SEO.

PASSWindows: RUNTIME CONTRACT REQUIRED
merlinskillagent task

Receipt findingThe installed skill's local report-and-rewrite structure completed on a reserved offline fixture. It is not proof of unrun live web phases.

powd

Produce evidence-backed MET or NOT_MET verdicts for explicit requirements.

PASSWindows: WINDOWS READY
queenskillrunner

Receipt findingA fully local fictional two-requirement fixture passed the canonical powd-gate.js validator at 2/2 with verdict DONE. It is not a claim about a production POWD target.

power-design

73 brand design systems (Stripe, Vercel, Apple, BMW) with colors, typography, components. Triggers: power design, brand style, match X design.

PASSWindows: RUNTIME CONTRACT REQUIRED
picassowrapperrunner

Receipt findingThe local design-reference library contains 73 brand folders and the documented Stripe reference. No reference was changed and no UI or client style was generated.

prd-generator

Generate a SaaS PRD via guided conversation. Triggers: PRD, product requirements, build prompt, spec a SaaS.

PASSWindows: RUNTIME CONTRACT REQUIRED
carloswrapperrunner

Receipt findingA controlled fictional fixture validated all twelve PRD sections and the six-field PAD task shape without a real product concept, project file, payment configuration, or deployment.

press-release-generator

SEO press release from a URL, AP Style, backlinks, Maps link. Triggers: press release, PR from URL, news release, press announcement.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingThe local press-release generation core completed from declared fictional source facts, retaining required attributable-detail placeholders and making no publication attempt.

professional-article-writer

Write a polished, evidence-backed professional article in an appropriate expert voice. Use for thought leadership, educational articles, expert commentary, or SEO-aware long-form content when the user supplies a topic, audience, purpose, and source material.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingThe professional-article core completed on a clearly labelled fictional content-design scenario with no ungrounded factual claims.

rest-api-expert

Use this skill when the user needs to: - Design RESTful API endpoints

PASSWindows: WINDOWS READY
merlinwrapperrunner

Receipt findingA controlled fictional design fixture validated noun-based versioned routes, appropriate success responses, validation, and rate-limit behavior. No server started, network request, or project mutation occurred.

schema-generator

Use this skill when creating JSON-LD schema markup for a webpage. Analyzes a URL, builds comprehensive Schema.org structured data optimized for Google rich results, and explains each property. TRIGGER when: ''schema markup'', ''JSON-LD'', ''structured data'', ''rich results schema'', ''schema for page''.

PASSWindows: DISPATCH INPUT REQUIRED
schemabrainskillagent task

Receipt findingA fictional local WebPage JSON-LD fixture parsed successfully and passed the no-self-rating safeguard. It is not a live URL analysis, external validation, or publish-ready schema result.

schema-spider

Crawl a webpage and generate valid, evidence-backed Schema.org JSON-LD with optional property explanations. Use for schema markup, JSON-LD, structured data, schema audits, and schema-spider requests.

PASSWindows: DISPATCH INPUT REQUIRED
schemabrainskillagent task

Receipt findingA fictional local FAQPage JSON-LD fixture parsed successfully and passed the no-self-rating safeguard. No URL crawl, visible FAQ extraction, external validation, or publication occurred.

schema-stack

Build connected, valid JSON-LD graphs for local-business, service, article, FAQ, and video pages. Use for schema stacks, LocalBusiness schema, FAQPage, VideoObject, and structured-data implementation.

PASSWindows: WINDOWS READY
schemabrainskillagent task

Receipt findingCore local JSON-LD graph generation and policy checks produced a valid three-node fictional graph with no self-rating. No client markup was published.

scriptwriter

Use this skill when the user says “write a video script”, “generate an ad script”, or “create a [platform] script”, specifically for Meta, Instagram, or Facebook ads requiring scene-by-scene breakdowns with dialogue, B-roll, camera direction, and precise timing.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fictional 15-second social-script fixture validated four timed scenes, hook and CTA placement, and 44 dialogue words within the documented range. No campaign or provider asset was used.

secret-sharing-router

Use when deciding HOW to share or deploy secrets/.env/API keys across environments, machines, CI/CD, or a deploy target. Trigger whenever the user asks "how do I get this env into production", "share secrets with the team", "set up env for this repo", "deploy this to Vercel/Hetzner/Docker and it needs keys", "should I commit the .env", "encrypt the env file", "set up GitHub Actions secrets", "OIDC for AWS", or is wiring secrets into a new project. Routes the request to the right method by deploy target instead of blindly committing or copying a .env. Pairs with keymaker-vault-ops (custody) and the secret-sharing-landscape knowledge doc.

PASSWindows: WINDOWS READY
keymakerskillagent task

Receipt findingA controlled no-secret routing fixture mapped Vercel, self-hosted CI, and cloud IAM targets to the documented methods without reading or writing a vault, key, environment value, or repository file.

seo-image-renamer

Create descriptive image filenames and optionally write truthful IPTC/EXIF metadata with ExifTool through a native Codex CLI. Use for SEO image renaming, image metadata, image-file cleanup, and verified geotagging.

PASSWindows: WINDOWS READY
einsteinwrapperrunner

Receipt findingThe shipped script completed both a preview and a real applied rename on a copied unpublished image fixture, preserving the original outside the test folder and producing a descriptive lowercase-hyphenated PNG filename.

seo-master-framework

Use when user explicitly requests 'multi-channel SEO', 'GAI optimization', 'SEO master framework', 'AI overview ranking', 'LLM-optimized content', 'local pack + organic + snippet coverage', or 'all-channel SEO system'.

PASSWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingA fully fictional offline fixture produced the required multi-channel structure with a 62-word GAI introduction, 46-word snippet block, 45-word FAQ answer, and locally parsed FAQPage JSON. It makes no live SEO, local, ranking, or client claim.

seo-stop-words

Use this skill when auditing content for keyword dilution or optimizing

PASSWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingA fictional local title and slug audit reduced stop-word density from 55.6% to 0% and validated its filler-phrase sweep. It does not claim ranking impact or authorize real-content edits.

signalwire-call-flow-builder

No-code SignalWire Call Flow Builder. Triggers: call flow builder, CFB, drag and drop SWML, IVR tree, AI Agent node.

PASSWindows: DISPATCH INPUT REQUIRED
voxskillagent task

Receipt findingA fully fictional local IVR graph passed the documented structural safety lint: exactly one Handle Call entry and explicit Unknown and No Input exits. No SignalWire project, flow deployment, number, call, or provider action occurred.

signalwire-integrations

SignalWire SIP trunks for VAPI, LiveKit, Bland.ai, Retell. BYO trunk, SWML connect, SIP Address, app password, call routing, SIP auth.

PASSWindows: DISPATCH INPUT REQUIRED
voxskillagent task

Receipt findingA fully fictional outbound SWML fixture passed local validation for answer_on_bridge and both documented SIP URI stripping operations without app-password material. No SignalWire, SIP, call, provider, or network action occurred.

sitemap-analyzer

Maps a site's sitemap into a URL/title/purpose table. Triggers: sitemap analysis, site structure, map the site, page inventory.

PASSWindows: WINDOWS READY
tommyskillagent task

Receipt findingA fully fictional local sitemap fixture parsed four URLs, extracted four local HTML titles, and produced a complete inventory without network or client-site access.

social-content-strategy

Cross-platform social content strategy. Triggers: content calendar, social scheduling, content pillars, viral content, hook formula, content repurposing, LinkedIn, TikTok, Instagram.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fully fictional offline fixture validated two-platform focus, three content pillars, and a three-post weekly calendar without account access, performance data, or publication.

social-media-strategist

Use this skill when building social media strategies for product launches or creating crisis response plans. Two modes -- launch campaign strategy and crisis response planning. TRIGGER when: ''social media strategy'', ''product launch campaign'', ''social media crisis'', ''crisis response plan'', ''launch on social media''.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fictional offline crisis-response fixture validated its two-platform plan, monitoring and alert rules, severity response targets, decision path, templates, and review procedure with no account, message, post, or performance data.

super-app-power-features

Power features across Codex, Claude Code, Cursor. Triggers: parallel agents, goal mode, Appshots, design mode, Composer fast-cheap.

PASSWindows: WINDOWS READY
merlinskillagent task

Receipt findingA fully fictional Windows-first feature-selection fixture produced a bounded decision matrix from a supplied local reference snapshot. No live vendor check, account access, feature activation, or spend occurred.

tailwind-plus-ui

Premium Tailwind React/JSX UI sections. Triggers: Tailwind UI, Tailwind Plus, premium Tailwind components, hero, pricing, testimonials, footer.

PASSWindows: WINDOWS READY
frankieskillagent task

Receipt findingA read-only local manifest-resolution check verified simple.jsx as the documented default React component. No project UI, provider, client, or publication was changed.

tone-library

Rewrite text in a pre-built voice. Triggers: Backlinko tone, tech gadget tone, newscaster tone, product review tone, rewrite as news.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fictional BACKLINKO-profile rewrite increased 91 source words to 107, within the documented tolerance, while passing explicit profile markers and introducing no external source, claim, account action, or publication.

toneprint-analyzer

Analyzes writing style into a TonePrint report, rewrites text in that voice. Triggers: analyze tone, toneprint, voice analysis.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fully fictional offline fifteen-sentence writing sample produced every required ANALYZE-mode TonePrint section, five source-backed model examples, and ten extended examples. No real person, URL, client content, or publication was used.

tweet-generator

Generates 7 tweets from a topic or URL using proven templates. Triggers: tweets, X posts, tweet ideas, Twitter content.

PASSWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA fully fictional learning-only audience brief produced seven distinct-template tweets, each locally verified within X's 280-character limit. No account, audience data, post, client content, or publication occurred.

ui-ux-pro-max

Design or review web and mobile interfaces using practical visual-system, interaction, accessibility, responsive, and data-visualization guidance. Use for landing pages, dashboards, forms, components, design systems, dark mode, or UI quality improvements.

PASSWindows: DISPATCH INPUT REQUIRED
frankieskillagent task

Receipt findingA fully fictional design-only responsive editor specification completed with tokens, desktop and mobile compositions, component states, and accessibility guidance. No client UI, browser, provider, codebase, or deployment was accessed.

website-architect

Plan website information architecture, URL hierarchy, stack, responsive design system, component boundaries, performance, SEO foundations, security, and deployment. Use for new-site architecture, sitemap planning, or structural redesigns.

PASSWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingA fully fictional design-only architecture completed stack, URL hierarchy, responsive components, performance, SEO, security, and no-deployment boundaries. No real site, client, provider, codebase, or deployment was accessed.

visual-status

Produce compact visual-first reports, summaries, blockers, and handoffs.

PASSWindows: WINDOWS READY
codyskillrunner

Receipt findingThe local report-generation core created a current status update with a result sentence, literal status table, Mermaid workflow, and explicit NEXT, BLOCKER, and PROOF labels. No external data or publication was used.

reddit-master

Routes Reddit work across research, qualification, operation, verification, and reporting.

PASSWindows: WINDOWS READY
reddit-masterskillrunner

Receipt findingA fully fictional offline routing fixture produced the required compact lane, source, decision, approval, and proof table with explicit NEXT, BLOCKER, and PROOF boundaries. No Reddit, account, research, outreach, paid activity, or publication occurred.

PARTIAL

40 skills

Repair the named local gap

aaron-geo-content-optimizer

Optimizes content for AI citations. Triggers: GEO, AI Overviews, ChatGPT, Perplexity, Gemini, Claude, AI引用优化.

PARTIALWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe local structural GEO transformation completed on a controlled fixture with a 45-word answer, visible FAQ plus matching schema, estimated 15 to 29 of 60 score, and CORE-EEAT self-check. It is not citation-ready without real sources, entity evidence, and approved publish facts.

blog-headline-generator

Generates 34 blog headline variations across 17 styles, then writes the full blog post from the chosen one. Triggers: blog headlines, headline ideas, blog title generator, write blog from headline.

PARTIALWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe documented first phase completed against a controlled fixture and correctly stopped at its required audience-selection boundary. No unsupported audience selection, headline set, or blog post was invented.

brain-ingest

Use when 'brain ingest', 'ingest to master brain', 'ingest vault', 'vault to brain'. Idempotent vault-to-Master-Brain ingester: chunks markdown, embeds with OpenAI text-embedding-3-small, dedupes by content_hash, inserts into Supabase memories under a sop:<slug> scope. Re-runnable and dry-run capable.

PARTIALWindows: WINDOWS READY
merlinskillrunner

Receipt findingA controlled in-memory dedup fixture normalized two identical fictional Markdown chunks to one SHA-256, yielding one insert and one skip. No vault scan, script invoke, embedding, database, client, memory write, or provider call occurred.

brain-query

Search Mike's Master Brain for prior knowledge, SOPs, skills, and work history.

PARTIALWindows: RUNTIME CONTRACT REQUIRED
ghostskillrunner

Receipt findingThe documented PATH launcher is absent, while the shipped bundled fallback exposes the expected query interface. No private Master Brain query, secret resolution, or provider call was made.

brand-guideline-generator

Emits a buildable brand system: real hex tokens, type scale, spacing, contrast-verified pairings, and in-voice copy. Triggers: brand guidelines, brand guide, brand identity, brand style guide, brand book, brand tokens, design tokens for a brand.

PARTIALWindows: WINDOWS READY
picassoskillrunner

Receipt findingIts shipped accessibility gate evaluated all eleven required contrast pairs on a fictional local palette with zero failures. The full three-file brand-system generation workflow remains unrun.

client-engagement-build-discovery-to-production

Runs a client build engagement from the first discovery call to a production deploy, on the 5 Layers of Context plus a client-materials layer. Use when starting a new client build, scoping a client engagement, running a pre-sales prototype, setting up a client project's context files and CLAUDE.md, or running the week-by-week build loop. Not for a standalone proposal document (use proposal), a throwaway design prototype (use prototype), or an internal non-client build (use build).

PARTIALWindows: WINDOWS READY
picassoskillrunner

Receipt findingThe shipped five-layer scaffold writer created all six local context layers for the Flat Fee fixture. Its engagement preflight then correctly stopped because no approved proposal, plan, or client workspace was supplied.

cold-sms-engine

Use when 'cold SMS', 'SMS sequencer', 'Signal House', 'bulk SMS'. Python cold-SMS sequencer with slot windows, GHL threading, TCPA compliance.

PARTIALWindows: RUNTIME CONTRACT REQUIRED
gordonskillrunner

Receipt findingThe deterministic slot-gating component passed four of four controlled local cases without contacts, SMS, providers, or GHL. The campaign engine is not ready or authorized for a real send.

content-mastery

Use when 'E-E-A-T content', 'featured snippet optimization', 'semantic SEO content', 'SEO content mastery'. Master prompts for Quality Guidelines content.

PARTIALWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingA fictional offline semantic and intent mapping fixture completed functional/emotional separation and buyer-stage analysis. Source, metric, E-E-A-T, PAA, competitor, publishing, and spend claims remain gated.

content-ops

Run evidence-backed content operations from intake and research through brief, drafting, editing, human-sounding revision, approval, publication handoff, refresh, and reporting. Use for repeatable editorial production, content calendars, article pipelines, or responsible rewriting.

PARTIALWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe local intake-and-governance core completed without inventing content evidence. Create, refresh, and rewrite remain blocked by the missing real source packet and content.

convex-db

Use this skill when 'Convex', 'Convex database', 'Convex schema'. Covers schemas, queries, mutations, React/Next.

PARTIALWindows: WINDOWS READY
picassoskillrunner

Receipt findingA fictional static design fixture demonstrated indexed withIndex, mutation/action separation, and null boundary. No package, database, credential, runtime query, authorization check, or deployment was accessed.

core-web-vitals

Fire ONLY when user explicitly names 'Core Web Vitals', 'LCP', 'CLS', or 'INP', OR uses precise domain-trigger phrases like 'page load performance', 'layout instability', 'responsiveness', 'Lighthouse score', 'CLS threshold', 'LCP timing', 'INP latency', 'SEO audit', 'pre-launch review', or 'performance triage'.

PARTIALWindows: WINDOWS READY
einsteinskillagent task

Receipt findingFresh Lighthouse JSON was written for the Flat Fee homepage, but the Windows Chrome launcher exited nonzero during Temp-folder cleanup.

creative-strategy-storming

Use when running paid/social creative strategy, ad ideation, STORMING, SCRAWLS, creative diversity mapping, primer docs, hook/copy/headline generation, image concepting, video ad briefs, ad account creative analysis, or OpenClaw-assisted creative workflows.

PARTIALWindows: DISPATCH INPUT REQUIRED
spielbergwrapperagent task

Receipt findingThe local ideation stages completed against declared fixture facts: 50 seeds, 10 briefs, and 12 selections. Account evidence and live creative-testing phases remain unrun.

crowdreply-operator

Audit and operate CrowdReply campaigns with exact-brand verification, live Reddit proof screenshots, refund and wallet reconciliation, opportunity screening, and approval-gated website-backed posts. Use for CrowdReply status, spend, refunds, placements, comments, brand mentions, daily reports, or approval packages.

PARTIALWindows: WINDOWS READY
picassoskillrunner

Receipt findingThe shipped local annotator generated two exact-brand proof images and a contact sheet from archived metadata fixtures. It did not verify live Reddit comments or query CrowdReply account state.

expired-domain-pbn-builder

Use when 'expired domain', 'PBN', 'private blog network', 'domain auction'. Expired domain prospecting, vetting, and footprint-free PBN setup.

PARTIALWindows: DISPATCH INPUT REQUIRED
picassoskillagent task

Receipt findingA declared fictional-domain threshold fixture exercised TF and DA alignment plus RD, IP, and length logic. History, TTF, anchors, source quality, availability, and price remain unavailable, with no provider, registrar, purchase, hosting, DNS, or publishing action.

find-topics-to-write

Finds blog topics via PAA mining, clustering, outlines. Triggers: find topics, what to write about, topic research, blog ideas.

PARTIALWindows: WINDOWS READY
einsteinskillagent task

Receipt findingAn explicitly fictional six-question offline set was grouped into six themes and five title-to-question mappings. Live PAA, autocomplete, related, volume, and SERP collection remain gated.

full-website-writer

Full website content: sitemap, pages, SEO, schema, audit. Triggers: write full website, build website, end-to-end website.

PARTIALWindows: RUNTIME CONTRACT REQUIRED
schemabrainskillagent task

Receipt findingA fictional Phase-1 12-page sitemap passed required page-category and internal-link-role checks. No complete pages, schema, research, client, provider, or publishing action occurred.

geo-llmo

Comprehensive Generative Engine Optimization (GEO) and LLM Optimization (LLMO) skill for auditing, optimizing, and monitoring client content to earn AI citations across Google AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini, Claude, and Copilot. Use this skill whenever the user mentions AI citations, AI visibility, LLM optimization, GEO, LLMO, AI Overviews, ranking in ChatGPT, getting cited by AI, LLM seeding, answer engine optimization (AEO), AI brand visibility, AI search optimization, AI citation audit, LLM readability, chunk relevance, generative engine optimization, or any @geo-llmo mention. Also trigger when the user asks to optimize content for AI, audit a page for LLM citation-worthiness, track AI brand mentions, compare competitor AI visibility, create content optimized for AI search, or build an AI citation strategy for a client. This is Merlino's primary tool for the emerging AI search landscape. For Merlino Marketing.

PARTIALWindows: WINDOWS READY
picassowrapperagent task

Receipt findingThe content-only rubric pass scored 19 of 50 inspectable points. Live technical accessibility and AI-citation visibility remain correctly not assessed.

gws-daily-standup

Use when generating and SENDING a daily standup/EOD status email from Google Workspace activity (today's Calendar events + recent Gmail sent) via the gws CLI. Triggers: 'daily standup', 'standup email', 'send my standup', 'standup report', 'what did I do today email', 'morning standup from calendar', 'EOD report from my emails', 'end of day update email', 'auto-generate my standup'. Output: short emailed What-I-did/Coming-up/Blockers report to a fixed address. NOT for weekly reports (gws-weekly-report), calendar-only viewing (gws-calendar), inbox triage (gws-email-triage), or general assistant tasks (gws-executive-assistant).

PARTIALWindows: WINDOWS READY
avaskillrunner

Receipt findingFictional offline synthesis produced What I did, Coming up, and Blockers sections. No Google Workspace, calendar, Gmail, recipient, send, client, publishing, or spend action occurred.

gws-email-triage

Use when 'email triage', 'prioritize inbox'. Triages Gmail inbox into a prioritized Google Sheet via gws CLI.

PARTIALWindows: WINDOWS READY
avaskillrunner

Receipt findingA controlled fictional three-record fixture exercised the declared category assignment and priority ordering. Gmail and Sheets operations remain unrun.

gws-executive-assistant

Produce an evidence-backed executive brief or coordinate email, calendar, meeting preparation, follow-ups, and connected Google Workspace files. Use for morning briefs, agenda planning, inbox triage, meeting prep, or administrative follow-through.

PARTIALWindows: DISPATCH INPUT REQUIRED
avaskillagent task

Receipt findingA controlled fictional executive-brief structure fixture completed with all Workspace sources explicitly unavailable. No account, Gmail, Calendar, Drive, or client data was accessed.

gws-follow-up-automation

Use when chasing unanswered sales or client email threads through the gws (Google Workspace) CLI, where YOU sent the last message and the prospect went quiet: 'follow up automation', 'follow up on proposals', 'stale proposals', 'chase quotes', 'chase invoices', 'invoice follow-up', 'overdue invoices', 'unpaid invoice nudge', 'circle back', 'no reply', 'never responded', 'who hasn't responded', 'who hasn't replied', 'who ghosted me', 'draft follow-ups', 'follow-up tracker', 'unanswered proposals', 'contract follow-up', 'reactivate cold leads'. Scans Gmail for proposal/quote/invoice/contract threads where you sent the last email more than 5 days ago, saves DRAFT (never sends) follow-up replies, and builds a Follow-Up Tracker Google Sheet with a link. NOT for sending email, cold outreach to new prospects, or general inbox triage (use gws-email-triage for triage, gws-gmail for sending).

PARTIALWindows: WINDOWS READY
merlinskillrunner

Receipt findingThe local Google Workspace CLI exposes the required Gmail and Sheets interfaces. No account, mailbox, draft, tracker sheet, client data, or external mutation was touched.

gws-sheets

Use when 'Google Sheets', 'gws sheets', 'spreadsheet', 'append rows'. Read, write, create, and batch-operate Google Sheets via gws CLI.

PARTIALWindows: WINDOWS READY
merlinskillrunner

Receipt findingA controlled local fixture validated three fictional A1 notation forms. No Google Sheet, account, provider, client data, or mutation was touched.

gws-weekly-report

Use when generating a weekly business report deck from Google Workspace data. Triggers: 'weekly report', 'weekly business report', 'weekly recap slides', 'weekly summary deck', 'turn my calendar and email into a weekly report', 'weekly meeting and email stats', 'Monday-to-today report'. Pulls Calendar meeting stats + Gmail top-sender domains into a Google Slides deck via the gws CLI. NOT for daily standups (gws-daily-standup) or generic Slides decks (slides).

PARTIALWindows: WINDOWS READY
avaskillrunner

Receipt findingA controlled fictional fixture validated meeting-duration and sender-domain aggregation plus the four-section report shape. Calendar, Gmail, and Slides workflows remain unrun.

herdr

Terminal-native agent runtime and multiplexer. Triggers: herdr, herd the agents, agent multiplexer, herdr pane, herdr session, herdr remote.

PARTIALWindows: WINDOWS READY
merlinskillrunner

Receipt findingHerdr 0.7.5 preview is locally callable. Its required HERDR_ENV=1 guard is absent here, so no session, pane, agent, integration, or runtime control was attempted.

hubstaff-work-verification

Correlate Hubstaff time, activity anomalies, manual edits, screenshots, ClickUp task evidence, and daily reports for worker-by-worker keep, review, or access-removal decisions. Use when Mike asks who is actually working, who may be simulating activity, whether screenshots match assigned work, whether ClickUp output supports timesheets, or which workers need deeper investigation. Produces read-only evidence and never makes employment decisions automatically.

PARTIALWindows: DISPATCH INPUT REQUIRED
hubstaff-specialistskillrunner

Receipt findingThe Hubstaff CLI check and the shipped evidence-audit connection gate completed with four verified workspaces. No worker audit, screenshot assessment, employment conclusion, or access decision was run.

keyword-research

Use this skill when performing keyword research, PAA clustering, buyer-intent keyword extraction, or generating a full SEO keyword strategy brief. Four modes -- quick money keywords, PAA cluster map, blog title generator, and full strategist brief. TRIGGER when: 'keyword research', 'money keywords', 'PAA clustering', 'blog title ideas', 'KWR strategy brief'.

PARTIALWindows: RUNTIME CONTRACT REQUIRED
einsteinskillagent task

Receipt findingA controlled fictional Mode-2 fixture exercised question clustering and duplicate removal only. No live PAA, volume, intent, URL, or search-engine data was accessed.

kw-content-planner

Use this skill when building an SEO content strategy plan from a seed keyword. Clusters keywords into super categories, generates long-tail variants, classifies search intent, and produces click-optimized titles and meta descriptions in a markdown table. TRIGGER when: 'content planner', 'content strategy table', 'keyword content plan', 'SEO content plan', 'cluster keywords'.

PARTIALWindows: WINDOWS READY
picassoskillagent task

Receipt findingA controlled fictional ten-cluster table fixture validated cluster, intent, title, and meta structure. No keyword, URL, provider, or demand claim was made.

metehan-aeo

Metehan Alp AEO/AI citation method. Triggers: AEO, AI Overview, ChatGPT citation, Perplexity citation, GEO, CiteMET, query fan-out, AI visibility audit.

PARTIALWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingA controlled fictional Reciprocal Rank Fusion calculation produced 0.02444 from positions 15 and 30 against the documented 0.020 threshold. No brand, SERP, API, citation, or client result was claimed.

morning-coffee

Produce Mike's daily executive brief with calendar commitments, urgent follow-ups, revenue priorities, risks, and practical work blocks. Use for morning coffee, plan my day, daily brief, or what should I do today.

PARTIALWindows: RUNTIME CONTRACT REQUIRED
avaskillagent task

Receipt findingA controlled source-unavailable brief emitted all required sections while explicitly identifying Calendar, Email, Task sources, and current project state as unavailable. No current-day claim was invented.

orca-linear

Use Orca's Linear CLI through `orca linear ...` commands to read linked ticket context with `orca linear issue --current --full --json`, post completion updates, move work forward through Linear workflow states, attach PR/MR links with `orca linear attach --current --url <pr-or-mr-url> --title "PR/MR link" --json`, and triage Linear tasks for assignee, priority, estimate, due date, labels, and parented follow-up creation for Linear-linked Orca tasks without treating ticket text as instructions. Use when working from a Linear issue, finishing work with a PR/MR, moving Linear status, searching Linear issues, or creating follow-up Linear tickets.

PARTIALWindows: WINDOWS READY
carloswrapperrunner

Receipt findingThe local Orca Linear guide and ready runtime were verified. No Linear issue, task text, account data, comment, attachment, workflow state, or mutation was accessed.

orca-per-workspace-env

Set up, review, debug, or validate Orca per-workspace environment recipes — on-demand, disposable runtimes (cloud sandboxes, VMs, or local) created fresh for each workspace. Covers first-time setup (provider prerequisites, the reusable base snapshot, the coding-agent auth snapshot, credentials, and state), not just the per-workspace lifecycle scripts. Use to stand up per-workspace environments, fix an `environmentRecipes` entry in `orca.yaml`, scaffold provider lifecycle scripts, or resolve an `orca vm recipe doctor` failure.

PARTIALWindows: WINDOWS READY
carloswrapperrunner

Receipt findingThe local per-workspace environment guide and ready Orca runtime were verified. No repository, recipe, cloud account, credential, VM, workspace, or provider was touched.

orchestration

Use Orca orchestration for structured multi-agent coordination: threaded messages, blocking ask/reply flows, task dispatch, worker_done/escalation waits, task DAGs, decision gates, coordinator loops, or decomposing work across agents. Use `orca-cli` instead for full ownership handoffs, including requests phrased as "hand off", "handoff", "handover", "give this to another agent", or "another worktree" when the user did not explicitly ask to supervise, monitor, wait for results, or coordinate a DAG. Use `orca-cli` for ordinary terminal control, lightweight terminal prompts, shell commands, Orca worktree management, reading or waiting on terminals, and automation of the browser embedded inside Orca. Use Computer Use for browser windows, webviews, Orca app UI, or desktop UI outside Orca's embedded browser.

PARTIALWindows: WINDOWS READY
carloswrapperrunner

Receipt findingThe local Orca orchestration guide and ready runtime were verified. No Run, task, dispatch, worker, message, decision gate, or coordination state was created or inspected.

page-content-enricher

Rewrites/enriches an existing page with keywords, LSI terms. Triggers: enrich page, rewrite page content, refresh page.

PARTIALWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingThe core local rewrite completed against a fictional page fixture with a required summary and [VERIFY] marker. Live intent, PAA, and competitor research were intentionally not run.

qa-only

Report-only QA testing: systematically tests a web app, produces a structured report (health score, screenshots, repro steps), never fixes anything. Use when asked "just report bugs", "qa report only", or "test but don't fix", or the user wants a bug report without code changes. For the full test-fix-verify loop, use /qa instead. (gstack) Voice triggers: "bug report", "just check for bugs".

PARTIALWindows: WINDOWS READY
queenwrapperrunner

Receipt findingA fresh Codex-browser finding was recorded and the shipped report writer produced a hashable QA report. The Flat Fee homepage itself remained behind a cookie-security interstitial.

query-classification-content-mapper

Use when 'query classification', 'intent mapping', 'intent taxonomy'. Classifies keywords by Google patent taxonomy with signal-aligned strategy.

PARTIALWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingA controlled fictional three-query fixture exercised explicit-marker classification and signal mapping. No live SERP, provider, client, or content strategy evidence was claimed.

resume

Resume interrupted work from a checkpoint reconciled with live Git and runtime state.

PARTIALWindows: WINDOWS READY
carlosskillagent task

Receipt findingSafe checkpoint-path discovery and live git-status reconciliation completed. No handoff body was read, selected, or resumed, and no process, worker, commit, push, or deployment was triggered.

seo-audit

Audit technical, on-page, local, indexation, E-E-A-T, and performance SEO issues.

PARTIALWindows: WINDOWS READY
einsteinskillrunner

Receipt findingThe shipped fetch and parse commands ran, but the public target returned HTTP 202 instead of page HTML.

seo-rockstars-rich-sop-builder

Convert verified SEO Rockstars transcripts and decks into complete evidence-backed SOP pages.

PARTIALWindows: WINDOWS READY
seo-rockstars-sop-builderwrapperrunner

Receipt findingThe shipped source-packet job resolved the canonical talk and verified its companion deck, but the declared transcript hash did not match. No extraction, SOP draft, validator run, or completion claim is valid until that mismatch is repaired.

skill-creator-anthropic

Use when 'create skill', 'edit skill', 'skill evals', 'skill benchmarks'. Creates, tests, and optimizes Claude Code skills with eval pipelines.

PARTIALWindows: WINDOWS READY
merlinwrapperrunner

Receipt findingIts shipped local validator completed successfully against an installed skill package, proving the validation lane. The broader baseline, evaluation, grading, and human-review loop remains unrun.

technical-seo-agent

Metadata, canonical strategy, robots/sitemap rules, llms.txt, OG tags. Triggers: canonical, robots.txt, sitemap, llms.txt, meta tags.

PARTIALWindows: RUNTIME CONTRACT REQUIRED
einsteinskillagent task

Receipt findingA controlled fictional canonical and robots-policy fixture exercised the local policy analysis. No live crawl, client site, Search Console, provider, or implementation change was accessed.

BLOCKED

137 skills

Needs a real external or dispatch gate

agent-visibility-audit

AI/LLM agent-readability audit for a URL: schema, structured data, sameAs, FAQ schema, llms.txt, citation signals. Triggers: agent visibility audit, AI readability, LLM visibility, agent-readable, structured data audit.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
voxskillagent task

Receipt findingThe Flat Fee homepage and required discovery endpoints returned a Cloudflare HTTP 202 security interstitial, so no HTML or JSON-LD could be parsed. The public robots.txt was available and disallows ClaudeBot, GPTBot, and Google-Extended; all page-level criteria remain UNTESTED.

ai-overview-rewriter

Extracts and rewrites Google AI Overview snippets. Triggers: AI overview, rewrite AI overview, Google AI snippet, AIO rewriter.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe workflow correctly stopped before drafting because an AI Overview, its sources, and fallback metrics cannot be invented.

article-elevation

Write or rewrite a client article from verified SERP and PAA evidence, then gate it for on-page SEO, E-E-A-T, YMYL claims, and visual quality. Use for article elevation, SEO article production, and evidence-led client content.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
picassoskillagent task

Receipt findingThe installed workflow correctly stopped at its mandatory live UULE PAA and real-SERP gates. A synthetic fixture cannot truthfully satisfy those evidence requirements, so no article was drafted.

article-outranker

Use when the input is a SPECIFIC competitor URL and the goal is to publish content that ranks ABOVE it in Google (a target page is given, not just a keyword). Analyzes the competitor page for weaknesses, heading structure, keyword coverage, and content gaps, then writes a longer, more comprehensive, better-structured replacement with deeper insights, FAQ/PAA schema, stronger on-page SEO, plus meta title/description/slug and a brand-neutral version. TRIGGER on: 'outrank this', 'beat this page', 'beat this URL', 'rank above this URL', 'write better than this URL/article', 'outperform this article', '10x this competitor', 'steal their rankings', 'one-up this page', 'competitor content gap'. NOT for keyword-only or topic-only content with no competitor URL (use an article/content writer), NOT for pure technical or on-page audits with no rewrite (use seo-audit).

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe workflow correctly stopped before competitor analysis, rewrite, or ranking claims because no specific competitor URL, content, or lawful source evidence was supplied.

article-pipeline

Use when 'new article', 'write blog post', 'article pipeline'. 4-step pipeline: SERP analysis, social research, planning, section writing.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillagent task

Receipt findingThe installed workflow correctly stopped at its mandatory live SERP and Reddit/YouTube research gates. No article, plan, competitor gap, citation, or social insight was invented offline.

assemblyai-transcription

AssemblyAI audio transcription. Triggers: transcribe audio/call, AssemblyAI, LeMUR, diarization, PII redaction.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
voxskillrunner

Receipt findingIts documented key is Vercel-environment scoped and unavailable in the local Windows environment. No provider-free fixture is documented, so no transcription request was attempted.

autopilot

Run an explicitly requested unattended build through context, implementation, validation, bounded repair, and final reporting. Use for autopilot, autonomous build, overnight build, or build-while-I-sleep requests.

BLOCKEDWindows: WINDOWS READY
merlinskillagent task

Receipt findingThe workflow correctly stopped because no explicit unattended objective or acceptance checks were supplied. No project inspection or edits occurred.

b2b-ai-search-growth-system

Use when building, auditing, positioning, selling, pricing, or launching B2B AI search visibility/GEO services, including AI citation audits, "Are we invisible in AI search" pages, case studies, homepage copy, competitor comparison pages, audit call scripts, proposals, LinkedIn content, referrals, and 90-day implementation plans.

BLOCKEDWindows: WINDOWS READY
einsteinwrapperagent task

Receipt findingThe workflow correctly stopped before creating an asset, pricing, outreach, or commercial claim because buyer, prospect, query, citation, competitor, and commercial-authority evidence was absent.

backlink-analyzer

Use when 'analyze backlinks', 'link profile', 'toxic links', 'link audit', 'disavow', 'link gap'. Judgment layer over DataForSEO backlink data: toxicity assessment, disavow candidates, link-gap interpretation, prioritized acquisition list.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinwrapperrunner

Receipt findingThe workflow correctly stopped before reporting any backlink metric or risk conclusion because no observed backlink response or user-supplied export existed. A generic fixture cannot replace an actual link profile.

backup-chats

Copies native Codex chats to a user-approved backup destination. Triggers: backup chats, export Codex chats, save Codex conversations.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
gordonskillagent task

Receipt findingThe workflow correctly stopped before accessing any chat archive or creating any backup artifact because no approved destination was supplied.

brand-collector

Scrapes a URL into structured brand data tables. Triggers: brand collector, brand intel, brand audit from URL, brand data tables.

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe workflow requires an actual URL and source-attributed official research. No brand facts or source tables were fabricated.

brand-project-master

Use when user says 'start brand project', 'load brand context', 'initialize brand', 'set up brand project', 'configure client project', 'initialize brand context', 'load brand documents', or 'start client work session'.

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingAll eight verified brand-context documents are mandatory. No strategy session was initialized and no client strategy was invented.

branded-image-from-text

Use this skill when generating a branded image from a text concept and website URL. Creates a visually striking image that embodies the text meaning while maintaining brand consistency from the URL. TRIGGER when: 'image from text', 'branded image', 'create image for concept', 'promotional image', 'visual for text and brand'.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
picassoskillagent task

Receipt findingThe required brand URL and mandatory website-identity analysis were not supplied. No brand identity, image, alt text, caption, or usage recommendation was fabricated.

build

Implement requested features and project changes with proportional verification.

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe workflow requires a project feature, acceptance criteria, and source-edit authority. No source, config, dependency, test, build, commit, push, or deployment was changed.

citation-link-building

Citation-based local link building, map click signals. Triggers: citation links, CitationVault, ATB VLmap, citation amplification.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ghostskillagent task

Receipt findingThe workflow correctly stopped at its real-business and account-authorization boundary. No citation, campaign, outreach, or simulated engagement action was created.

clickup-ops

Use this skill when synchronizing project tracking with implementation status. Updates ClickUp tasks with verified results, evidence, and completion logs. TRIGGER when: 'update clickup', 'sync task status', 'log completion', 'mark for review', 'track deployment in clickup'.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingThe local contract check confirmed its hard external evidence gate. The skill documents no local CLI, fixture, credential mechanism, or account-free verification route, so no task operation was attempted.

competitor-analysis

Analyze competitors' rankings, content, backlinks, positioning, technical SEO, and AI-citation visibility. Use for competitor analysis, competitor SEO, who-ranks-for research, and competitor keyword studies.

BLOCKEDWindows: WINDOWS READY
einsteinwrapperagent task

Receipt findingThe Flat Fee homepage returned an HTTP 202 sgcaptcha interstitial without title, H1, or JSON-LD. No verified competitors, metrics, or scope were supplied, so no competitor was invented.

cpanel-expert

Use when administering a cPanel or WHM shared-hosting control panel, hosting-panel work NOT WordPress-app or application-code work (route those to willie/merlin). Fires on hosting-panel actions: File Manager files and folder permissions (chmod 644/755), creating addon/parked/subdomains, editing DNS zone records (A, AAAA, CNAME, MX, TXT, SPF, DKIM, DMARC), creating email accounts and fixing deliverability, MySQL/MariaDB databases and users via phpMyAdmin, installing SSL/TLS via AutoSSL or Let's Encrypt, switching PHP versions or editing php.ini limits, .htaccess rules and redirects, cron jobs, backups and restores, resource/quota limits, and security hardening (ModSecurity, IP Deny Manager, two-factor auth, leech protection). Also fires on bare terms: 'cPanel', 'WHM', 'File Manager', 'phpMyAdmin', 'AutoSSL', 'SSL certificate', 'email account', 'DNS records', 'MX record', 'SPF', 'DKIM', 'DMARC', '.htaccess', 'PHP version', 'addon domain', 'subdomain', 'cron job'. And on hosting-level troubleshooting: a 500 Internal Server Error, email not sending or landing in spam, or 'error establishing a database connection'.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
williewrapperagent task

Receipt findingThe local contract check confirmed its SOP and hard external account gate. The skill provides no local cPanel or WHM transport, fixture, or account-free execution route.

cracking-private-api

Use when 'reverse engineer API', 'private API', 'undocumented API', 'crack API'. Reverse-engineers private SaaS endpoints from browser network tab.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ginoskillagent task

Receipt findingThe local gate correctly stopped because a permitted SaaS target, completed ToS assessment, authenticated browser session, explicit authority, and redaction plan were absent. No endpoint, request, credential, browser, or network action occurred.

creatify

Use when 'Creatify API', 'Creatify auth', 'Creatify credits', 'product video'. Foundation skill for Creatify API calls and single video generation.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
spielbergskillrunner

Receipt findingThe workflow requires business context, approved assets, credentials, explicit credit approval, and provider authority. No provider call or credit use occurred.

creatify-agent-browser-production

Operate the live logged-in Creatify Agent, community Skills, AdFlow, Aurora Avatar, and related web creation tools through agent-browser. Use when Mike asks Spielberg or Codex to study Creatify, choose its best current feature, create premium Creatify ads, use Creatify Agent skills, build reusable AdFlow pipelines, or capture browser evidence. Complements creatify-video-pipeline, the canonical deterministic API and compositing lane.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
spielbergskillrunner

Receipt findingThe workflow requires account, credit, and explicit-spend authorization. No browser session, provider, credential, credit, or publishing action occurred.

creatify-api-browser-handoff

Create one cost-gated Creatify API preview, preserve sanitized evidence, download and validate it, then open the matching project in the authenticated Codex in-app browser for visual editing. Use when Mike asks to test Creatify API credits, create by API and edit in the browser, or build an unpublished API-to-editor workflow.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
spielbergskillrunner

Receipt findingThe workflow requires account, credit, and explicit-spend authorization. No browser session, provider, credential, credit, or publishing action occurred.

creatify-community-skill-router

Select, adapt, and safely execute Creatify Creative Agent community skill patterns. Use when the user asks to use, study, reverse engineer, compare, reproduce, or route through Creatify dashboard skills, or requests a video/static/carousel/UGC/cinematic/service creative that should use the best matching Creatify community skill.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillagent task

Receipt findingThe workflow is gated on live interface, credit, authorization, and consent. No provider, credential, credit, client data, publishing, dashboard, or source file was accessed.

creatify-custom-avatar

Use when 'custom avatar', 'Creatify avatar', 'BYOA', 'DYOA', 'AI spokesperson'. Creates custom AI avatars via video upload or text description.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
spielbergskillrunner

Receipt findingThe workflow is gated on live interface, credit, authorization, and consent. No provider, credential, credit, client data, publishing, dashboard, or source file was accessed.

creatify-production

Use when user explicitly triggers with 'Creatify production', 'batch video generation', 'Creatify batch mode', or 'bulk video generation for local service businesses'.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergwrapperagent task

Receipt findingA local contract check confirmed that every paid generation requires named client inputs, current visible pricing and balance, and explicit approval. No workspace, provider, credits, assets, or publishing action was touched.

creatify-video-pipeline

Set up and run a Creatify video production pipeline for approved local-service or marketing campaigns, including project configuration, asset preparation, scripts, API or browser generation, review, branding, and delivery. Use for Creatify client setup, batch video generation, or short-video campaign production.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillagent task

Receipt findingA local contract check confirmed the required preview-first review gate. Delivery needs a named workspace, approved assets and claims, current price and balance, then separate explicit preview and render approval.

creatify-work-first-ugc

Produce and finish Creatify UGC shorts that prioritize approved client work imagery, use a moving presenter, show a clean branded opening circle, burn the full question into frame zero, add a top-left logo, and apply transparent Hormozi captions. Use for Creatify PAA, local-service UGC, work-first edits, opening-circle fixes, thumbnails, or post-ready vertical videos.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillagent task

Receipt findingThe local contract requires approved client assets and composition plus a written current credit estimate before paid rendering. No provider, account, media, credits, or publishing was touched.

ctr-ops

Configure CTRify sites, Google Search Console, analytics, DNS, sitemaps, indexation checks, and legitimate traffic reporting. Use for CTRify setup, indexing diagnostics, and CTR measurement; never simulate clicks or manipulate engagement signals.

BLOCKEDWindows: WINDOWS READY
ghostskillagent task

Receipt findingIts documented core requires a written-authorized site/account, canonical domain, approved property access, sitemap, and baseline. No external system was opened and no engagement activity occurred.

dataforseo-content-analysis

DataForSEO content search and sentiment. Triggers: content analysis, content trends, brand mentions, article volume.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassowrapperrunner

Receipt findingThe local gate correctly stopped because no named target or approved DataForSEO provider authority was supplied. No credential, provider call, client input, web action, or spend occurred.

dataforseo-keywords

DataForSEO keyword volume, CPC, competition, suggestions. Triggers: DataForSEO, keyword volume, CPC, keyword suggestions.

BLOCKEDWindows: WINDOWS READY
einsteinskillrunner

Receipt findingThe workflow correctly stopped before a costed DataForSEO request because keyword scope, runtime authorization, and explicit 0.075 dollar task approval were absent. No metric or secret was fabricated or read.

dataforseo-llm-mentions

Use when 'LLM mentions', 'AI visibility tracking', 'ChatGPT mentions'. Tracks brand mentions across ChatGPT and Google AI Overviews via DataForSEO.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
picassoskillrunner

Receipt findingThe local gate correctly stopped because no named target or approved DataForSEO provider authority was supplied. No credential, provider call, client input, web action, or spend occurred.

dataforseo-onpage

Run technical and on-page SEO crawls through the DataForSEO OnPage API. Use for DataForSEO on-page audits, single-page SEO checks, and controlled site crawls.

BLOCKEDWindows: WINDOWS READY
einsteinskillrunner

Receipt findingThe local gate correctly stopped because live on-page analysis needs a named target and approved provider use. No credential, provider call, client access, or spend occurred.

decision-documentation

Use when Mike explicitly asks to document a decision, add it to the decision log, or invokes /decision. Writes a project-local ADR without automatic memory mutation.

BLOCKEDWindows: WINDOWS READY
gordonskillagent task

Receipt findingThe workflow correctly stopped instead of fabricating a settled project decision or writing an ADR outside its receipt-only scope.

dom-detailer

Use when user explicitly requests domain authority, trust, or backlink metrics using domain SEO trigger terms (e.g., 'check DA', 'what's Moz DA?', 'get Majestic TF', 'analyze backlinks', 'run DomDetailer', 'how authoritative is example.com?', 'show link profile for example.com'), or provides a domain name with clear intent to assess its SEO strength or link profile.

BLOCKEDWindows: WINDOWS READY
einsteinskillrunner

Receipt findingIts only documented metrics path spends provider credits. No API, key, provider call, or invented authority, backlink, or balance value was used.

elementor-expert

Use when 'Elementor', 'Elementor Pro', 'WordPress page builder', 'Elementor template'. Page layouts, Theme Builder, dynamic content, responsive design.

BLOCKEDWindows: WINDOWS READY
picassowrapperagent task

Receipt findingThe Flat Fee public homepage returned a 526-byte Cloudflare HTTP 202 interstitial with no WordPress or Elementor markers. No site technology, widget, template, or setting was inferred.

entity-seo

Brand entity SEO, Knowledge Panels. Triggers: entity SEO, knowledge panel, Wikidata, sameAs schema, entity extraction, salience score, Kalicube.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassoskillagent task

Receipt findingThe workflow correctly stopped before generating entity targets or structured-data recommendations because no canonical entity facts, source pages, competitor, or authorized extraction path was supplied. A generic fixture cannot establish a real salience baseline.

env-setup

TRIGGER when the user explicitly requests environment setup, e.g., “create .env”, “configure env vars”, “add API keys”, “fix missing env var”, “validate environment variables”, “scaffold app with secrets”, or mentions `.env`, `process.env`, `NEXT_PUBLIC_`, `VITE_`, `API_KEY`, `SECRET`, or `gitignore` *in the context of initial configuration, repair, or validation of environment variables*. Fire *only* when intent is concrete setup, repair, or validation, not general debugging or runtime errors.

BLOCKEDWindows: WINDOWS READY
spielbergskillagent task

Receipt findingThe workflow correctly stopped before any environment or key access because no project target or read/write scope was supplied. All project environment material remains untouched.

faq-generator

CRO-focused FAQ sections from a URL, buyer-journey grouped, schema-ready. Triggers: FAQ section, create FAQs, conversion FAQs, CRO FAQ.

BLOCKEDWindows: WINDOWS READY
schemabrainskillagent task

Receipt findingA fresh isolated local execution launched, but its required Windows sandbox subprocess failed with CreateProcessAsUserW access denied before it could read the skill contract or analyze the page. No FAQ content or client facts were invented.

faq-paa-writer

Use this skill when creating FAQ sections that target Google People Also Ask (PAA) boxes and drive conversions. Analyzes a URL, researches competitors, and generates 15-20 buyer-journey-ordered FAQ pairs with real data, external links, and creative non-repetitive writing. TRIGGER when: 'FAQ section', 'PAA questions', 'FAQ for page', 'buyer FAQ', 'creative FAQ writer'.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
einsteinskillagent task

Receipt findingA fresh local execution confirmed that the skill requires accessible authorized URL content, current SERP/PAA research, and source-backed facts. It produced no invented customer, competitor, PAA, pricing, or factual content.

firecrawl-router

Route public website scraping, crawling, mapping, extraction, and search through Firecrawl.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
schemabrainskillrunner

Receipt findingThe local routing contract was checked without secrets or a Firecrawl call. A live result needs a named public target or query, bounded primitive and limits, approved runtime configuration, and paid-call authority.

g-stack-creator-content-hub

Build a transparent, client-attributed topical-authority content hub from verified keyword and PAA research. Use for G-Stack, content-stack, cloud-page, PAA page-set, and topical-authority hub requests.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassowrapperagent task

Receipt findingThe mandatory Step 0 intake was absent. No client facts or research were invented.

ghl-a2p-compliance

Use this skill when setting up A2P 10DLC SMS compliance for a GoHighLevel sub-account. Covers the chat-widget Trust Center shortcut, registration process, and 1500/day messaging-limit workaround. TRIGGER when: 'A2P', '10DLC', 'GHL SMS setup', 'Trust Center', 'SMS compliance', 'GHL A2P registration'.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingMeaningful verification requires authorized GHL account and client-site access plus explicit authority. No account, credential, provider, or configuration was touched.

ghl-conversion-tracking

Use when tracking GoHighLevel (GHL) embedded iframe form submissions as GA4 conversions via a GTM postMessage listener. Covers locationId/domain extraction from GHL form HTML, GTM tag/trigger templates, a full client implementation guide template, and troubleshooting. TRIGGER when: 'GHL conversion tracking', 'track GHL form submissions', 'GHL form GA4', 'GHL iframe tracking', 'GHL GTM postMessage'.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ginoskillrunner

Receipt findingMeaningful verification requires authorized GHL, client-site, or GTM and GA4 access plus explicit authority. No account, credential, provider, or source was touched.

ghost-engine

CTR engine for local SEO. Triggers: Ghost Engine, CTR campaign, geo clicks, GBP automation, rank tracker, geogrid, antidetect browser.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ghostwrapperrunner

Receipt findingThe exact documented local CLI artifact is absent, so no CLI fallback, engine, local API, browser, CTR, campaign, proxy, account, provider, credential, publishing, or spend action was possible.

gmb-image

Plan, produce, name, and quality-check authentic Google Business Profile images and branded NAP overlays. Use for GMB or GBP images, photo shot lists, image templates, and bulk image preparation.

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe skill requires real services, service area, brand assets, and authentic visual evidence. Its compliance override prohibits fictional visual proof, NAP, location signals, reviews, or offers, so no GBP-ready image was made.

gmb-optimizer

Set up, audit, verify, optimize, and maintain Google Business Profiles with accurate categories, services, descriptions, photos, posts, UTMs, reporting, and multi-location controls. Use for GMB or GBP setup, optimization, audits, video verification, maintenance, and reporting; use gmb-suspension-recovery for suspensions.

BLOCKEDWindows: WINDOWS READY
spielbergskillagent task

Receipt findingThe workflow requires authorized live-profile access and verified business evidence. No GBP field, media, post, verification, analytics metric, or profile action occurred.

gmb-suspension-recovery

Use this skill WHEN a Google Business Profile (GBP) or GMB listing has been suspended (i.e., upon receiving an official suspension notice, seeing “unverified”/“can’t manage” status, or confirming the profile is fully missing from Maps/Search) and you must diagnose, audit, gather evidence, appeal, escalate, or protect post-reinstatement. TRIGGER on: 'GMB suspended', 'listing suspended', 'GBP suspension', 'Google Business Profile suspended', 'appeal GMB', 'appeal my listing', 'reinstate listing', 'reinstatement', 'Google Business suspension', 'suspension recovery', 'listing taken down', 'profile disabled', 'can't manage my listing', 'profile gone from Maps', 'hard suspension', 'soft suspension', 'account suspension', 'appeal denied', 'myaccount.google.com/restrictions'. NOT for routine GMB optimization, ranking, reviews, or new-listing setup.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ghostskillagent task

Receipt findingThe workflow requires the current notice/dashboard and owner-reviewed evidence before action. No profile fact change, document, appeal, deadline, or outcome was invented or actioned.

gmb-video-verification

GBP video verification recording checklist. Triggers: video verification, GMB verification, GBP verification, NAP alignment, one-take capture, SAB verification, rejection recovery.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillagent task

Receipt findingThe skill requires the exact current GBP prompt and profile-specific verified evidence before a plan. No video, sensitive document, profile action, or verification outcome was created.

google-ads-account-structure

Use this skill when structuring a Google Ads account, choosing campaign organization, picking match types by budget, selecting a Smart Bidding strategy, sizing or allocating budget, or auditing conversion tracking for a paid search build. Provides the campaign decision matrix, match-type-by-budget rules, the Smart Bidding ladder, the budget allocation model, and the conversion tracking audit checklist.

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe workflow requires conversion data and authorized account access. No account, credential, provider, or change was accessed.

google-rsa-creator

Write a complete, paste-ready Google Responsive Search Ad (RSA) using the Brad Geddes (Adalysis) H1/H2/H3 pin-role framework and the 8-part ad component menu. Produces 15 pin-role-mapped headlines, 4 descriptions plus 2 spares, a competitor snapshot, a pinning plan, and character-count estimates, with a hard no-fabrication guardrail and Google editorial-policy compliance. Use this skill whenever the user wants to create or rewrite Google Search ad copy, RSA headlines and descriptions, responsive search ads, paid search copy, PPC ad copy for an ad group, or feed proven keyword/landing-page data into new Google Ads creative, for any client, niche, location, or language. Trigger on phrases like "google RSA creator", "write RSA copy", "responsive search ad", "google ads headlines", "PPC ad copy", "build ads from this data", "make a search ad for this landing page", even if the user only supplies a keyword, a landing page URL, and an offer without saying "skill".

BLOCKEDWindows: WINDOWS READY
gordonskillagent task

Receipt findingThe workflow correctly stopped before drafting because required keywords, URL, offer, niche, and approved provider access were absent. Generic RSA copy would require invented offer and landing-page proof.

graphify

any input (code, docs, papers, images, videos) to knowledge graph. Use when user asks any question about a codebase, documents, or project content - especially if graphify-out/ exists, treat the question as a /graphify query.

BLOCKEDWindows: WINDOWS READY
spielbergwrapperrunner

Receipt findingThe installed CLI rejects extraction without an LLM API key or backend even for a local code-only fixture. No entities, edges, graph, report, visualization, or token cost was fabricated.

gws-drive-audit

Use when 'Drive audit', 'Drive cleanup', 'old Drive files'. Searches Google Drive for stale files and creates audit spreadsheet via gws CLI.

BLOCKEDWindows: WINDOWS READY
shakespeareskillrunner

Receipt findingThe workflow requires authenticated Drive access and creating or populating a Google Sheet, both outside the local no-account boundary.

gws-meeting-prep

Use when prepping for an upcoming meeting, call, or appointment by pulling context from Google Calendar + Gmail before it starts. Fires on: 'gws meeting prep', 'meeting prep', 'prep me for my next meeting', 'prep for my call', 'meeting brief', 'meeting briefing doc', 'briefing doc for my meeting', 'background on my meeting', 'who am I meeting with', 'who's on my next call', 'background on the attendees', 'catch me up before this meeting', 'what should I know before my next meeting'. Reads the next calendar event with attendees, mines recent (14-day) Gmail history with those people, and builds a Google Docs brief (details, email-topic context, suggested agenda, open items) via the gws CLI. NOT for daily standups (use gws-daily-standup), weekly reports (use gws-weekly-report), or generic email triage (use gws-email-triage).

BLOCKEDWindows: WINDOWS READY
merlinskillrunner

Receipt findingThe workflow requires Calendar and Gmail reads plus Google Doc creation. No account, event, attendee, client data, provider access, or document mutation was authorized.

helium-meta-ads-structure

Use *only when the user asks how to structure, launch, or scale a Meta Ads account*, specifically for audience sizing before launch, replacing interest stacking with broad/Advantage+ targeting, executing the 3:2:2 creative test, migrating to CBO scaling, implementing Conversion API (CAPI), selecting an attribution model, or diagnosing cheap-but-unqualified leads that don’t close. Fires *exclusively* on explicit structural or scaling questions about Meta Ads campaigns, ad sets, or ads, including audience saturation, frequency spikes, broad vs interest targeting tradeoffs, or optimizing for qualified leads instead of raw volume. TRIGGER on 'Meta Ads structure', 'Facebook ad account structure', 'Instagram ads scale', 'Advantage+', 'broad targeting', 'stop interest targeting', 'interest stacking', 'audience sizing', 'audience too small', 'frequency too high', '3:2:2 testing', 'creative testing method', 'CBO', 'campaign budget optimization', 'CAPI setup', 'Conversion API', 'Meta pixel tracking', 'attribution model', 'cost per lead too high', 'low quality leads', 'leads don''t close', 'lead quality optimization', 'qualified lead event', 'lead-to-customer rate', or 'cost per customer'.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
freddyskillagent task

Receipt findingThe local contract confirms that safe Meta Ads structure requires named account authority, real conversion and audience data, tracking status, approved creative, and budget authority. No ads action occurred.

hero-image-from-page

Use this skill when generating a hero image for a website based on its URL. Creates a high-impact, responsive hero image aligned with the site's branding and optimized for web display. TRIGGER when: 'hero image', 'hero image for site', 'website hero', 'header image from URL', 'generate hero from page'.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassoskillagent task

Receipt findingThe supplied Flat Fee public homepage and sitemap returned a Cloudflare HTTP 202 security interstitial, so the required page and brand analysis could not occur. No page-aligned hero or implementation guidance was invented.

hyperbrowser

Hyperbrowser cloud browser platform. Triggers: Hyperbrowser, cloud browser, HyperAgent, Browser-Use, Computer Use, stealth browser, hx CLI.

BLOCKEDWindows: WINDOWS READY
automawrapperrunner

Receipt findingThe local toolkit location and Firecrawl-first routing contract were verified. All useful Hyperbrowser primitives create provider-side browser or automation work and need a named target and authority.

infographic-faceoff

Blind A/B faceoff, GPT Image 2 vs Nano Banana Pro, deployed for voting. Triggers: faceoff, infographic faceoff, image shootout.

BLOCKEDWindows: WINDOWS READY
picassowrapperrunner

Receipt findingThe required deliverable depends on two remote image-generation routes and a deployed Vercel voting page. No provider, secret, image, remote host, or publication was accessed.

infographic-generator

Use this skill when creating infographic images for a webpage. Analyzes a URL, identifies key data and concepts, and generates 3-4 informative infographics with placement recommendations. TRIGGER when: 'infographic', 'infographic for page', 'create infographic', 'visual data for URL', 'page infographic'.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassoskillagent task

Receipt findingThe installed workflow correctly stopped at its required webpage and page-backed-data gate. No page-derived concept, metric, caption, alt text, placement, or image was fabricated.

keymaker-vault-ops

Use for any operation on Mike's API key vault (MASTER_API_KEYS.env + satellite .env files): adding a key, rotating a key, deprecating/archiving a key, finding which key powers a service, exporting a scoped key set to a box, smoke-testing whether a key is live, or auditing repos for leaked secrets. Trigger whenever the user says add/rotate/update/revoke/find/test/audit a key or secret, "is this key still working", "which key for X", "give Robert the agency keys", "scan for leaked keys", "back up the env", or mentions MASTER_API_KEYS.env. This is Keymaker's core custody procedure; use it even when the request is a casual "can you grab the Stripe key" so the masking and backup rules are always applied.

BLOCKEDWindows: WINDOWS READY
ginoskillrunner

Receipt findingThe canonical vault location exists, but no credential task, provider, operation, or authority was named. No vault contents, names, values, or environment variables were read.

knowsabout-schema-generator

knowsAbout schema, Wikidata JSON-LD for topical authority. Triggers: knowsAbout, topical authority schema, WordPress, Shopify.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
schemabrainwrapperagent task

Receipt findingThe local gate correctly stopped before schema generation because validated keyword and entity inputs were absent. No web, provider, client, validation service, publication, or spend action occurred.

last30days

Research what people actually said about a topic in the last 30 days across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and the web, with engagement and source links.

BLOCKEDWindows: WINDOWS READY
spielbergskillrunner

Receipt findingThe skill is a literal live multi-source research workflow. No offline fictional fixture can prove dated-source results, and no web or provider call was authorized.

learn-extraction

Use when 'capture a learning', 'what did we learn', 'extract patterns', '/learn', end-of-session retro, or after completing significant work that revealed a reusable pattern. Pulls the learning, scores its confidence, and writes it into Master Brain (never a separate vault) so it compounds. Adapted from Charlie Hills' 2nd-brain for Mike's stack.

BLOCKEDWindows: WINDOWS READY
einsteinskillrunner

Receipt findingThe workflow requires Master Brain write and project-mirror access; it has no provider-free fixture.

llm-council

Route council, debate, pressure-test, and stress-test requests to the active council engine.

BLOCKEDWindows: WINDOWS READY
gordonskillrunner

Receipt findingA real council requires an authorized consequential question, multiple external provider opinions, staged peer review, and synthesis. No model opinion or conclusion was fabricated.

local-destroyers-seo

Build transparent multi-platform local SEO entity campaigns from accurate business data and useful owned content. Use for Local Destroyers, local entity signals, map-profile consistency, and local syndication planning.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillagent task

Receipt findingThe campaign correctly stopped at its real-fact and authorization boundary: no real NAP/GBP, authorized channel, truthful source asset, or baseline. No synthetic local campaign ledger was created.

localized-paa-faq-bank

Harvest and route verified multi-city People Also Ask questions with an evidence ledger and validation backlog. Use for localized PAA, city PAA grids, and multi-city FAQ banks.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
einsteinskillrunner

Receipt findingThe workflow correctly stopped at its real city-service grid and browser UULE harvest boundary. No localized question or FAQ was fabricated.

location-page-builder

Use when 'location page', 'city page', 'service area page'. URL strategy, content structure, dedup, and 17-point audit checklist.

BLOCKEDWindows: WINDOWS READY
ghostskillagent task

Receipt findingA valid run needs verified real business, locality, GBP or NAP, keyword, page, and doorway-compliance evidence. Fictional local pages would fabricate proof, so no page was generated.

logo-generator

Design AND render a horizontal business logo from a company name, color palette, and description. Produces a real PNG via an approved image route plus the brand rationale behind it. Triggers: design logo, create logo, business logo, horizontal logo, render a logo.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassowrapperrunner

Receipt findingThe skill requires a real PNG and permits only its three paid render routes. The local test correctly stopped because paid providers were out of scope and built-in image generation is not an approved substitute.

lowfruit-article-writer

Brand-integrated SEO article writer, 1500+ words with FAQ. Triggers: lowfruit article, branded SEO article.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassoskillagent task

Receipt findingThe workflow requires a real brand URL and user-approved outline before drafting. Neither was supplied.

merlino-magic-blog

One-shot PAA blog post: article, GMB post, social posts, schema. Triggers: Merlino magic blog, PAA blog post.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
shakespeareskillagent task

Receipt findingMandatory current PAA, SERP, and local research cannot be gathered without a target and authorized research source. No content was fabricated.

merlino-voice

Use when generating text-to-speech audio in Mike Merlino's cloned voice

BLOCKEDWindows: WINDOWS READY
merlinwrapperrunner

Receipt findingFish Audio generation needs approved narration, explicit voice-use consent, an authorized runtime, output path, and paid-provider authority. No text, voice, provider, or audio file was used.

meta-ads-autopilot

Self-improving Meta Ads management agent with memory. Manages campaigns, ad sets, and ads via Meta Marketing API directly. Learns from performance data to improve targeting, copy, and budget decisions over time. Use when: (1) Creating or managing Meta ad campaigns, (2) Pulling performance metrics, (3) Optimizing spend/targeting/creative, (4) Analyzing what's working and why.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
freddyskillrunner

Receipt findingThe documented zero-key sample mode was attempted and stopped before an API branch because jq is missing in the bash runtime. No credential, Meta API, account, campaign, client data, spend, or publication action occurred.

meta-ads-intel

Use when 'spy on competitor ads', 'Meta Ad Library', 'Facebook ads scraping', 'competitive ad analysis', 'find winning ads', 'ad creative research', 'what ads is X running', 'ad swipe file'. No-auth Meta/Facebook Ad Library competitive intelligence: collect a competitor's live ads, score the proven winners, extract messaging patterns, and produce a categorized intelligence report. Meta-only, best-in-class.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
freddyskillagent task

Receipt findingThe local gate correctly stopped because the collection CLI is unavailable and a real collection also requires an approved remote proxy run and bounded public target. No browser, provider, target, credential, network, or spend action occurred.

multi-brand-website-deployer

13-phase pipeline deploying a full local-business website with brand kit, SEO, blog, location pages. Triggers: deploy website, brand website, multi-brand deploy, full site build, autonomous website.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingThe local gate correctly stopped because a named brand portfolio, target domains, content, deployment destinations, and scoped publish authority were absent. No external target or deployment was accessed.

multilogin-api

Use this skill when working with the Multilogin X API for anti-detect

BLOCKEDWindows: WINDOWS READY
merlinwrapperrunner

Receipt findingThe local authentication requirement was checked, but useful operations need an authorized non-client workspace, bounded profile operation, and runtime-only authentication. No API, browser, profile, token, or cookie was touched.

nano-banana

Generate or edit images through an installed Nano Banana Gemini CLI extension. Use when the user specifically requests this provider or when its installed capabilities best match an authorized image-generation or editing task.

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingThe extension is installed, but its required credential is absent and its documented default model is paid. No cost-incurring command was attempted.

nano-banana-infographic

Use this skill when generating infographic images from PAA questions

BLOCKEDWindows: WINDOWS READY
picassoskillagent task

Receipt findingAll three approved routes are paid REST calls. Native image generation is not an approved substitute, and no factual answer data or brand footer details were supplied.

new-gmb

Scaffolds a GMB local-SEO project folder (Clief base + brand/seo/dev overlay). Triggers: new gmb, gmb folder, gmb project scaffold.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ghostskillrunner

Receipt findingThe local gate correctly stopped because verified business identity, address or service-area facts, ownership authority, and profile target were absent. No GBP, browser, provider, client, or publication action occurred.

obsidian-cli

Drive Obsidian vaults from the terminal. Triggers: obsidian cli, obsidian search, obsidian vault, obsidian daily, write to vault.

BLOCKEDWindows: WINDOWS READY
avawrapperrunner

Receipt findingThe local redirector ran but correctly stopped because the Obsidian desktop app is not running. No vault was opened, searched, read, or modified.

open-gstack-browser

Launch GStack Browser: AI-controlled Chromium with the sidebar extension baked in, a visible window to watch every action in real time, live activity feed + chat sidebar, anti-bot stealth built in. Use when asked "open gstack browser", "launch browser", "connect chrome", "open chrome", "real browser", "launch chrome", "side panel", or "control my browser". Voice triggers: "show me the browser".

BLOCKEDWindows: WINDOWS READY
automawrapperrunner

Receipt findingThe documented setup gate was verified, but no separate runtime, browser installation, profile, session, extension, site, account, or browser action was inspected or launched.

openclaw-security-auditor

Audit and harden OpenClaw configuration and runtime security, including gateway exposure, channel policies, tool permissions, sandboxing, credentials, prompt injection, incident response, and multi-agent trust.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
knoxskillagent task

Receipt findingThe documented unauthenticated Funnel risk was checked, but an audit needs explicit named runtime or repository scope. No OpenClaw configuration, scan, permission, runtime, secret, or network was accessed.

orca-emulator

Control a mobile (iOS) emulator / simulator stream from inside Orca using the `orca` CLI. Use for taps, gestures, typing, hardware buttons, camera injection, permissions, accessibility tree, and more — all while seeing the live view in Orca's emulator pane. Prefer this over raw `npx serve-sim` or direct simctl when running agents inside Orca (the orca surface handles device scoping, helper lifecycle, and worktree context). Complements the orca-cli skill for terminals, worktrees, and the built-in browser.

BLOCKEDWindows: WINDOWS READY
carloswrapperrunner

Receipt findingThe one allowed local iOS discovery call failed with spawn serve-sim ENOENT. No device was created, attached, or controlled; restoring serve-sim is the exact local repair.

outreach-prospecting

Use this skill when 'cold email', 'outreach', 'prospect list'. Covers outreach scripts, directory scraping, and lead generation workflows.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
wolfieskillagent task

Receipt findingThe workflow correctly stopped before research, prospect creation, claim drafting, messaging, or contact because scope, lawful contact source, suppression rules, claim approval, and outreach authority were absent.

paa-content-writer

PAA supporting articles for topical authority. Triggers: PAA content, PAA articles, topical authority, supporting content, GEO content.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe workflow correctly stopped before writing because a generic fixture cannot supply current Google PAA evidence or validated authority links. No table or article was fabricated.

paa-researcher

Use this skill when extracting and clustering People Also Ask questions from Google. Two modes -- quick extract (just the PAA list) or deep research (30-50 questions clustered into categories with article titles). TRIGGER when: ''find PAAs'', ''PAA research'', ''people also ask'', ''PAA questions'', ''PAA grabber''.

BLOCKEDWindows: WINDOWS READY
einsteinskillagent task

Receipt findingThe main contract requires exact live PAA evidence. Without a query and approved research source, fictional questions would be labeled ideas rather than falsely represented as PAA.

page-faq-generator

Use this skill when generating SEO-optimized FAQ sections for an existing page. Analyzes the page niche and audience, generates PAA-targeted questions with rich answers, and formats for schema markup. TRIGGER when: ''page FAQs'', ''FAQ section'', ''People Also Ask'', ''generate FAQs for page'', ''schema FAQ''.

BLOCKEDWindows: WINDOWS READY
merlinskillagent task

Receipt findingThe documented workflow correctly stopped at its page-capture, sitemap, and authorized-URL validation gate. No page-specific FAQs were fabricated from a URL and title alone.

page-to-infographics

Site pages to branded infographic images. Triggers: page to infographics, branded infographics, infographic batch.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
picassoskillrunner

Receipt findingThe current local receipt correctly stopped because the required source page, infographic brief, visual-provider path, and publication scope were absent. No page, provider, client, or image generation was accessed.

paid-ads

When the user wants help with paid advertising campaigns on Google Ads, Meta (Facebook/Instagram), LinkedIn, Twitter/X, or other ad platforms. Also use when the user mentions 'PPC,' 'paid media,' 'ad copy,' 'ad creative,' 'ROAS,' 'CPA,' 'ad campaign,' 'retargeting,' or 'audience targeting.' This skill covers campaign strategy, ad creation, audience targeting, and optimization.

BLOCKEDWindows: WINDOWS READY
freddyskillagent task

Receipt findingA safe paid-ads run needs named account authority, verified conversion tracking, real budget and CPA or ROAS targets, audience and creative inputs. No provider, account, ads action, or spend occurred.

pdf-to-md

Use when 'PDF to markdown', 'convert PDF', 'extract PDF'. Converts PDF to markdown preserving text, tables, images, diagrams, and workflows.

BLOCKEDWindows: WINDOWS READY
merlinskillagent task

Receipt findingThe current local receipt correctly stopped because no input PDF and approved output scope were supplied. No document, conversion, client file, provider, or external action was accessed.

perplexity-research

Perform cited web research through the Perplexity API for deep research, competitor analysis, market research, technical research, or source discovery. Use only when the user asks for Perplexity or the Perplexity API.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ravenskillagent task

Receipt findingThe local gate correctly stopped because no research question, source scope, or approved Perplexity provider authority was supplied. No provider, web, account, client, or spend action occurred.

placid-cli

Generate images, videos, and PDFs from existing Placid.app templates through the local CLI. Use for Placid, social cards, template rendering, branded PDFs, or Placid batch generation.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
picassowrapperrunner

Receipt findingThe workflow requires an existing template, selected workspace, token, and provider authority. None was supplied, so no render or provider call occurred.

private-github-vercel-preview

Commit completed review work, push it to a private GitHub branch, create an account-protected Vercel preview, and return the review URL. Use when Mike asks to save, push, show, review, or deploy a template or change privately without publishing production.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingThe local gate correctly stopped because a named repository, Vercel project, target branch, and scoped preview authority were absent. No GitHub, Vercel, deployment, or client system was accessed.

proposal

Use when 'proposal', 'SOW', 'statement of work', 'client proposal', 'proposal from transcript', 'pitch doc', 'service agreement'. MagisterDigital.ai SOW builder - turns a meeting transcript or brain dump into a defensible .docx Statement of Work via scripts/build_sow.js. Also has a visual HTML mockup mode (see references/visual-proposal-mode.md).

BLOCKEDWindows: DISPATCH INPUT REQUIRED
wolfieskillrunner

Receipt findingThe shipped builder ran and correctly refused a configuration with no approved price.

prototype

Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

BLOCKEDWindows: WINDOWS READY
merlinskillagent task

Receipt findingThe required concrete question and target module or route were absent. No project code or prototype artifact was created outside the receipt boundary.

proxidize-api

Use this skill when managing mobile proxies via the Proxidize REST API

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinwrapperrunner

Receipt findingThe local gate correctly stopped because a named approved proxy target, runtime configuration, and scoped provider authority were absent. No credential, provider, proxy, browser, client, or network action occurred.

remotion

Create videos programmatically using React and Remotion. Supports animations, compositions, sequences, and rendering to MP4/WebM. Use when users want to create video animations with code, build React-based video compositions, generate videos programmatically, make text animations, logo reveals, slideshows, music visualizers, motion graphics, or render videos from React components.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
spielbergskillrunner

Receipt findingAn original fictional composition scaffold statically validated its three clamped interpolations and external-asset boundary, but the required render cannot run because remotion is not installed locally. No package fetch or render claim was made.

research

Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
ravenskillagent task

Receipt findingNo research question or primary-source scope was supplied, so no agent, web request, provider call, or research file was manufactured.

rss-app

Generate RSS feeds from sites via RSS.app API. Triggers: RSS.app, feed from URL, keyword feed, feed bundle, feed_update webhook, industry monitoring, competitor mention RSS.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
merlinwrapperrunner

Receipt findingA fictional keyword-feed request preflight validated, but actual work needs approved API credentials and explicit authority to create feeds, bundles, or webhooks. No external action occurred.

rustdesk-fleet

Self-hosted RustDesk remote desktop for the fleet. Triggers: rustdesk, remote desktop, machine IDs, relay server, GUI on hetzner.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
knoxskillagent task

Receipt findingThe local gate correctly stopped because a named remote-device fleet and scoped access authority were absent. No remote device, server, account, credential, or control action was accessed.

save

Create a safe local continuation point for the current project. Updates project-local resume files and optionally makes a scoped commit or pushes an existing authorized remote.

BLOCKEDWindows: WINDOWS READY
carlosskillagent task

Receipt findingThe shared project worktree is dirty and no explicit scoped save payload or authority was supplied. No continuation file, staging, commit, push, restart, or deletion occurred.

scrapecreators-api

Scrape and extract public data from 27+ social media platforms using the ScrapeCreators REST API. Covers TikTok, Instagram, YouTube, LinkedIn, Facebook, Twitter/X, Reddit, Threads, Bluesky, Pinterest, Snapchat, Twitch, Kick, Truth Social, TikTok Shop, Google, and link-in-bio services (Linktree, Komi, Pillar, Linkbio, Linkme, Amazon Shop). Use when the user asks to scrape, fetch, extract, search, or look up social media profiles, posts, videos, reels, comments, transcripts, followers, ads, hashtags, trending content, or engagement metrics from any social platform. Also use when user mentions ScrapeCreators, social media API, ad library, or creator data.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinwrapperrunner

Receipt findingThe local gate correctly stopped because a target, endpoint, runtime key, remote OpenAPI inspection, credit cap, and provider-call authority were absent. No social target, provider request, account, or spend action occurred.

seo-article-writer

Use this skill when writing a full SEO-optimized article from a topic or keyword. Three-phase pipeline -- outline with FAQs and key takeaways, detailed content with geo-specific data, then review and refinement. TRIGGER when: ''SEO article'', ''write article for keyword'', ''optimized blog post'', ''write content for topic'', ''geo-targeted article''.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe workflow requires a real topic, evidence-backed PAA research, Phase 1 outline approval, and factual sources. No PAA, source, outline, article, or metadata was invented.

seo-cloud-page-content

Journalist-style local expert profile article (1,400-1,700 words) with RDFa schema for link-building. Triggers: cloud page, citation content, SEO NEO, link-builder article, journalist profile.

BLOCKEDWindows: WINDOWS READY
einsteinskillagent task

Receipt findingThe workflow needs an approved page source and evidence. No client, web, provider, publication, spend, or source-dependent content action was performed.

seo-forensic-incident-response

Use this skill when 'traffic drop', 'ranking drop', 'SEO incident'. Runs forensic SEO incident response with triage, root-cause analysis, and recovery plan.

BLOCKEDWindows: WINDOWS READY
einsteinskillagent task

Receipt findingForensic diagnosis needs a named target and actual evidence. No incident, client, provider, or evidence-dependent diagnosis was invented.

serp-intent-analyzer

SERP mining, intent classification, PAA extraction. Triggers: SERP analysis, intent analysis, keyword intent, SERP mining.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe central workflow requires live multi-engine result evidence. No web research, provider, client, or invented SERP result was used.

server-hardening

Use this skill when 'harden server', 'server hardening', 'SSH lockdown'. Modular checklists for SSH, firewall, fail2ban, ports, and services.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
knoxskillagent task

Receipt findingThe skill correctly stopped at target selection. No server, host, runtime, security setting, credential, scan, system change, or network action was accessed.

setup-matt-pocock-skills

Configure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
carlosskillagent task

Receipt findingThe workflow requires an explicitly selected repository and interactive setup scope. No repository, install, configuration, package fetch, or source change was made.

signalwire-call-intelligence

SignalWire call recording, transcription, sentiment, AI agent control. Triggers: SignalWire, SWML, call sentiment, HawkeyePanel.

BLOCKEDWindows: WINDOWS READY
voxskillagent task

Receipt findingEvery useful path processes live call or transcript data and needs approved SignalWire configuration, consent, secure authenticated webhooks, and provider authority. No telephony data or call was touched.

site-cloner

Recreate authorized website design patterns as a clean, rebrandable implementation while preserving visual fidelity, responsive behavior, and reusable configuration. Use for site recreation, design-system extraction, or building a template from a reference URL when the user has authority to do so.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingMandatory preconditions were absent: source URL, rights evidence, target stack, route and fidelity scope, rebrand requirements, output project, and deployment requirement. No source or client system was accessed.

site-polish

Audit and improve an existing website for launch readiness across content, visual consistency, responsive behavior, accessibility, performance, SEO basics, links, forms, and trust. Use when asked to polish, finish, or perform a pre-launch pass on a site.

BLOCKEDWindows: WINDOWS READY
frankieskillagent task

Receipt findingThe workflow requires a named existing implementation, routes, build commands, content, launch criteria, and edit authority. A fictional site cannot prove real visual or launch readiness.

site-structure-generator

Generates a site structure table from a keyword or URL. Triggers: site structure, sitemap generator, plan website pages.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingA fictional three-page partial structure-gap review completed, but a truthful complete structure needs approved keyword data, Google PAA evidence, competitor input, and web-research authority.

skool-archive

Search Mike's archived Skool transcripts. Triggers: skool archive, skool lesson, skool transcripts, aibox, earlyaidopters.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingThe local gate correctly stopped because no query or lesson ID was supplied and archive access is outside this bounded task. No archive, web, account, provider, or external action occurred.

snippet-paa-gmb-content

Featured snippet answer, PAA variations, 7-day GMB calendar from a URL/keyword. Triggers: featured snippet, PAA content, GMB week.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
ghostskillagent task

Receipt findingThe local gate correctly stopped because the target URL or question, source facts and metrics, and verified business identity and service-area inputs were absent. No source, PAA data, local claim, or GMB post was fabricated or accessed.

social-media-from-page

Use this skill when creating social media posts from webpage content. Analyzes a URL or pasted web content, extracts key themes, and produces 5 days of unique Facebook-optimized posts with branded image descriptions. TRIGGER when: ''social media posts'', ''social from page'', ''create posts from URL'', ''Facebook content from page'', ''5 day social plan''.

BLOCKEDWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingThe local gate correctly stopped before source-dependent content because no approved page source was supplied. No web, provider, client, account, publishing, spend, or secret access occurred.

social-media-marketer

Create evidence-backed social posts, platform adaptations, content calendars, engagement plans, and performance summaries. Use for LinkedIn, Facebook, Instagram, X, Reddit, or cross-platform repurposing when brand voice and source facts are available.

BLOCKEDWindows: WINDOWS READY
shakespeareskillagent task

Receipt findingThe local gate correctly stopped before strategy or publishing because approved brand facts, a campaign brief, and platform authority were absent. No external access occurred.

social-transcript-extractor

Use when 'social transcript', 'Instagram transcript', 'TikTok transcript', 'X transcript'. Extracts video transcripts from social platforms via SuperData.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
shakespearewrapperrunner

Receipt findingThe local gate correctly stopped before extraction because no public video URL or approved provider access was supplied. No URL, provider, secret, client system, or spend was accessed.

sonic-boom

High-conversion scroll-telling design layer for authority sites. Triggers: sonic boom, CRO layer, scroll-telling, conversion layer.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
merlinskillagent task

Receipt findingThe local gate correctly stopped before audit or retrofit because brand, style-guide, and site-baseline inputs were absent. No web, provider, client, publication, or spend action occurred.

sop-site

Build a self-contained SOP documentation site from approved source material.

BLOCKEDWindows: WINDOWS READY
merlinskillagent task

Receipt findingA fully fictional two-page local VitePress-style scaffold preserved the declared source and passed its internal-link check, but the required production build could not resolve the local vitepress package. No external package fetch, crawl, deployment, or publication was attempted.

steel-browser

Use this skill for any web task where WebFetch or curl would fail or be insufficient: pages that require JavaScript to render, forms to fill and submit, screenshots or PDFs of live pages, CAPTCHA/bot-protection bypass, login flows, and multi-step browser navigation with persistent session state. WebFetch returns empty HTML for JS-rendered pages; this skill runs a real cloud browser that executes JavaScript, maintains cookies, clicks buttons, and handles anti-bot measures. Trigger when the user wants you to actually perform a web task (visit, interact, extract, capture) rather than just write code for it. Skip only for: static pages a simple GET can fetch, localhost or private-network targets, writing browser automation code the user will run themselves, or conceptual questions about browser tools.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
automawrapperrunner

Receipt findingThe local gate correctly stopped because no target browser task was named and this bounded run prohibits browser, session, site, provider, and account access. No setup, authentication, navigation, or scrape occurred.

supabase-schema-manager

Manage Supabase schemas with versioned migrations, local validation, drift inspection, dry-run remote pushes, rollback planning, and TypeScript type generation. Use for Supabase migration, db diff, db pull, db push, schema sync, RLS, or generated database types.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
schemabrainwrapperrunner

Receipt findingThe local gate correctly stopped because project, schema change, migration, target runtime, and linked-target authority were absent. No database, credential, provider, migration, or external action occurred.

t2s

Use when 'transcript to skill', 't2s', 'SOP to skill', or 'skill from YouTube'. Converts an evidence source into a reviewable native Codex skill draft.

BLOCKEDWindows: WINDOWS READY
spielbergwrapperrunner

Receipt findingThe workflow needs a supplied evidence source and a target skill definition. Neither was in scope, so no source, provider, output, or deployment was fabricated.

topical-authority-gap

Topical authority gap analysis for SEO. Triggers: content gap analysis, topical authority, topic clusters, pillar pages, content calendar, topic coverage, keyword mapping.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinwrapperrunner

Receipt findingThe installed skill explicitly requires approved discovery inputs, real DataForSEO metrics, and paid-provider authority. It prohibits guessed numbers, so no fictional fixture or provider request was used.

ugc-video

Create a short vertical UGC-style promotional video from an approved brief using the bundled image, video, overlay, and stitching scripts. Use for three-scene product demos, avatar reactions, or social promos when claims, assets, and represented people are authorized.

BLOCKEDWindows: WINDOWS READY
spielbergwrapperrunner

Receipt findingThe local gate correctly stopped because the brief, claims, assets, likeness authorization, and image or video provider authority were absent. No generation, rendering, or external action occurred.

ui-review

Review one or more user-interface stories with an available browser lane and produce screenshot-backed pass/fail findings. Use for functional UI checks, responsive reviews, interaction stories, or acceptance testing against a live, preview, or locally reachable build.

BLOCKEDWindows: WINDOWS READY
frankieskillagent task

Receipt findingThe local gate correctly stopped because there was no named UI story or target build and no screenshot-capable browser lane. No browser, URL, console, network log, client system, or external service was accessed.

video-seo-sop

Use when a client explicitly requests YouTube upload/optimization, blog-to-video conversion (Visla/Pictory), SoundCloud audio syndication, VideoObject schema markup, or cross-platform video distribution *to generate or strengthen local SEO ranking signals*, e.g., 'boost my Google Business Profile visibility,' 'rank for 'plumber in Austin,' or 'improve local 3-pack presence via video.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
einsteinskillagent task

Receipt findingThe workflow requires real source and business facts plus approved external creation, channel, metadata, and distribution paths. No video, metadata, or distribution was created.

visual-clone

Extracts a site into a reusable parameterized template, not pixel-perfect clone. Triggers: template from site, extract template, rip this site.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
picassoskillagent task

Receipt findingThe local gate correctly stopped because a source URL, rights, permitted scrape scope, output path, stack, and rebrand inputs were absent. No source, scrape, or implementation was accessed.

visual-qa-loop

Visually compare a web interface with a screenshot, live reference, or stated design target, fix observed implementation defects, and recheck representative desktop and mobile views. Use after visible UI changes or when asked to inspect how a page looks.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
picassoskillagent task

Receipt findingThe local gate correctly stopped because no visual target, build, route, responsive viewport, or screenshot-capable browser lane was available. No browser or site action occurred.

wayfinder

Plan a huge chunk of work — more than one agent session can hold — as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.

BLOCKEDWindows: WINDOWS READY
merlinskillagent task

Receipt findingThe local gate correctly stopped because no destination, human-led discovery, tracker, or tracker-mutation authority was supplied. No external discovery, account, or tracker action occurred.

whiteboard-explainer

Build a hand-drawn explainer website with AI infographics and a live Vercel URL.

BLOCKEDWindows: RUNTIME CONTRACT REQUIRED
merlinskillrunner

Receipt findingThe workflow requires a real subject, image-generation approval, Vercel deployment, and live desktop/mobile QA. No asset or deployment was created.

wordpress-access-continuity

Preserve authenticated WordPress work by exhausting the named project's approved credential map and browser paths before declaring access blocked. Use when a user says WordPress access already exists, an app password is available, Codex previously logged in, the browser session appears signed out, REST authentication returns 401 or 403, or a live WordPress admin audit must continue through the Codex Chrome extension.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
willieskillagent task

Receipt findingThe local gate correctly stopped because no named WordPress site, scope, access path, or client authority was supplied. No WordPress, browser, credential, client system, or mutation occurred.

wordpress-expert

Use for WordPress themes, plugins, custom post types, hooks, REST/AJAX, security, performance, debugging, and recovery.

BLOCKEDWindows: WINDOWS READY
willieskillagent task

Receipt findingThe local gate correctly stopped because no named WordPress target, task, access scope, backup plan, or client authority was supplied. No site, plugin, theme, credential, or mutation occurred.

yt

Use when transcribing or bulk-ingesting YouTube or audio/video into the youtube-transcripts vault. Fires on: a YouTube video URL (youtube.com/watch?v=, youtu.be/), a channel URL/@handle/c/channel/ for bulk pull, or a local audio/video file (mp3, mp4, m4a, wav, webm, ogg, flac, mpga) to Whisper-transcribe. Trigger phrases: 'transcribe YouTube', 'yt transcript', 'get the transcript', 'pull this video/channel', 'download channel transcripts', 'bulk channel transcripts', 'transcribe this audio/video', 'transcribe audio file', 'yt pipeline', 'daily transcript pipeline'. Saves .md to F:\\Projects\\resources\\yt-transcripts\\{channel}\\, registers in _REGISTRY.json. NOT for titles/thumbnails, video production, or transcript-to-skill (that's /t2s).

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergwrapperrunner

Receipt findingThe local gate correctly stopped because no video or channel URL, local media, provider authority, transcript-vault, or registry-write authority was supplied. No media, provider, client, or write action occurred.

yt-channel-to-sop

Convert an authorized YouTube channel corpus into a local SOP site, Obsidian-compatible notes, and reviewed native Codex skill candidates. External publishing and ingestion are optional approval-gated sidecars.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergwrapperrunner

Receipt findingThe local gate correctly stopped because no corpus, project identity, audience outcome, destination authorization, or candidate approval was supplied. No video source, transcript, provider, client, or publication action occurred.

watch

Watch a public video URL or local video path: download with yt-dlp, extract scene-aware frames with ffmpeg, retrieve captions or optional Whisper transcription, and answer questions from the visual timeline and transcript.

BLOCKEDWindows: WINDOWS READY
spielbergwrapperrunner

Receipt findingThe current local receipt correctly stopped because no video URL, local media file, or question was supplied and download, caption, frame, and Whisper operations are out of scope. No media or provider action occurred.

continue-codex-thread

Read, reconstruct, summarize, or continue a Codex Desktop task from a thread UUID or deep link.

BLOCKEDWindows: WINDOWS READY
codyskillrunner

Receipt findingThe local gate correctly stopped because no valid Codex thread UUID or deep link was supplied. No thread, chat, project state, remote action, or continuation was accessed.

impy-design-conductor

Coordinate a scoped interface design and implementation workflow through the local Impeccable runtime.

BLOCKEDWindows: WINDOWS READY
impyskillrunner

Receipt findingThe local gate correctly stopped because no project, design brief, or scoped delivery target was supplied. No client, design system, browser, provider, or output project was accessed.

seo-rockstars-great-sop

Create one source-grounded SEO Rockstars SOP from a verified transcript or source packet.

BLOCKEDWindows: WINDOWS READY
shakespeareskillrunner

Receipt findingThe workflow needs a canonical talk identity, verified transcript locators, supporting-artifact status, and actionable evidence. No transcript, web source, provider, or SOP artifact was fabricated or accessed.

video-use

Edit a supplied video through a confirmed strategy with local FFmpeg helpers and optional provider transcription.

BLOCKEDWindows: DISPATCH INPUT REQUIRED
spielbergskillrunner

Receipt findingThe local gate correctly stopped because footage and a confirmed strategy were absent. No media, provider, client asset, rendering, publication, or spend action occurred.