Development

Diagnose why an existing implementation differs from reference evidence using AE/SSIM, pixel, section, computed-style, and transition diffs. Use for post-implementation mismatch or repair guidance; not baseline capture or full clone generation.

Clone or recreate a live website URL, page, or section as React + Tailwind with extracted DOM, CSS, assets, responsive layout, motion, and interactions. Use for live-URL implementation or fidelity repair; not capture-only or diff-only requests.

Capture baseline evidence from a live website URL: screenshots, scroll states, hover/click behavior, parallax, timers, and transition recordings. Use when reference artifacts are needed before implementation or comparison; not to build or diagnose a clone.

Guided interactive setup for scout's search engines and fetching tools. Covers Context7, Jina Reader, Exa, and Playwright. Triggers when setting up scout, configuring API keys, adding search engines, troubleshooting fetch failures, or improving search quality.

Web search and content fetching. Find relevant sources and read web pages with precision.

Fetch web page content from URLs. Read articles, documentation, and web pages with privacy-aware tool selection. URL取得、ページ読み込み、コンテンツ取得、Webページ、記事取得、ドキュメント取得、fetch、read page、get content

Claude implements, Codex reviews, repeat until clean

Codex reviews, Claude filters and fixes, repeat until clean

Run a Crucible review pipeline on the current project. The Profiler reads the project, interviews the user, and casts a 4–8 persona review committee from a 23-persona library. Peers review at the code level, departments hunt for gaps, leadership grades alignment, and the Aggregator synthesizes a final score and verdict. Output is a live terminal stream plus a fully-detailed markdown report at .review/reports/<id>.md.

List past Crucible reviews stored in .review/reports/ and optionally show the full markdown of one. Use to see how the project's review verdicts have evolved over time.

Refresh the project's .review/aims.md interactively without running a full Crucible review. Use when the project's stated goals have shifted between reviews and you want to update aims before the next /crucible run.

Audit which MCP server tools your Claude Code sessions actually invoke. Use when the user wants to find dead MCP servers (configured but never called), hallucinated MCP tool calls, or per-project MCP usage. Triggers on "which MCP servers don't I use?", "clean up MCP config", "remove unused MCP servers", "audit MCP tools". Runs locally; reads ~/.claude session JSONL logs and ~/.claude.json. No network calls.

Use when starting a PR review and the session file needs to be created or refreshed. Not user-facing — invoked by the orchestrator only.

Use when walking a user through code changes file-by-file during a PR review. Not for automated or silent reviews.

Use when running a PR review in --auto mode, to post analyzer findings as a pending GitHub review without an interactive walkthrough. Not user-facing — invoked by the orchestrator only.

Ingest/curate side of the video intelligence plugin. Use whenever the user wants to modify or build the corpus: scan YouTube channels for new videos and generate mind maps via Gemini; transcribe a video (URL or local MP4); run the full mindmap+transcript+concepts pipeline on a local video; extract and normalize concepts into the taxonomy; rebuild the LanceDB hybrid-search index from existing transcripts; clean up title-rotation duplicates; prune YouTube Shorts that polluted the corpus; manage...

Query a pre-built video corpus (mindmaps, transcripts, concept taxonomy, hybrid search index) produced by the video-intel skill. Use whenever the user wants to: find videos about a topic; look up what a creator said about something; retrieve evidence or quotes from transcripts; browse concepts in the library; synthesize a cross-creator brief ("nugget") grounded in indexed evidence; ask about corpus status (last scan, video counts); summarize a specific video that is already in the corpus; dec...

Translate YouTube videos and rich English transcripts into Bosnian/Croatian/Serbian (BCS) subtitles via Gemini. Use this skill whenever the user wants to: translate a YouTube video to Bosnian, Croatian, Serbian, or Serbo-Croatian; produce BCS captions or subtitles for a video; download just the English SRT from a YouTube video with no translation; or translate a context-rich transcript (with on-screen content and speaker labels) into BCS while preserving that context. Trigger phrases include ...

Expert on Adobe Acrobat PDF forms - creating, modifying, and filling form fields, understanding field types and properties, JavaScript actions, form validation, and integration with Adobe Acrobat Sign (e-sign). Use when the user asks about PDF form fields, AcroForm fields, e-signature workflows, or programmatic PDF form manipulation.

Show the current state of the project index: staleness, module coverage, hook health, and recent sync activity. Use to diagnose index issues.

Create specifications through parallel analysis

Rebuild the index for a specific module or the entire project. Accepts --module flag for targeted rebuild and --m flag for custom instructions.

Implement through pattern learning

Generate a complete project index with per-module README files for efficient code navigation. Use when starting work on a new project or after major restructuring. Run once per project.