Index Youtube
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill is mostly coherent for indexing YouTube transcripts, but it recommends using Chrome browser cookies for authenticated YouTube access without declaring or bounding that credential use.
Review before installing. The transcript indexing workflow is largely purpose-aligned, but do not allow the `--cookies-from-browser chrome` option unless you intentionally want yt-dlp to use your signed-in browser session. Verify solograph/yt-dlp package sources before running install commands, and keep downloads scoped to known channels and the documented docs/youtube output path.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If run, the downloader may act as the user's signed-in browser session when contacting YouTube/Google.
This tells yt-dlp to use the user's Chrome browser session/cookies for authenticated requests, while the registry declares no credentials and the artifact does not bound which cookies are used or how authenticated account access is controlled.
Reduce `-n` limit, add `--sleep-interval 2` to yt-dlp commands, or use `--cookies-from-browser chrome` for authenticated access.
Avoid the cookie option unless necessary; require explicit user approval first, prefer unauthenticated downloads or a dedicated browser/profile, and document the account/cookie scope.
Running these commands installs or executes external package code on the local machine.
The setup path fetches or executes unpinned third-party packages. This is aligned with the skill's purpose, but provenance and versions are not fixed in the artifact.
pip install solograph # or uvx solograph
Verify the package source, pin trusted versions where possible, and run setup commands only after user approval.
The skill can create local files and run download/search commands as part of indexing.
The skill uses broad local tools and shell commands to download transcript files and write an index. This is expected for the purpose, but the user should control channels, limits, and output paths.
allowed-tools: Bash, Read, Glob, Grep, Write, AskUserQuestion ... yt-dlp --write-auto-sub --sub-lang en --skip-download ... -o "docs/youtube/%(channel)s/%(title)s.%(ext)s"
Use explicit channel handles, keep limits small, use --dry-run when unsure, and keep outputs in the documented docs/youtube directory.
Indexed transcript content may be reused in future searches or summaries, even if the video content is inaccurate or adversarial.
The skill persists external YouTube transcript content and summaries for later search. This is the intended function, but untrusted video content can later influence retrieval or analysis.
Index YouTube video transcripts into a searchable knowledge base ... Write a summary index to `docs/youtube/index.md`
Treat indexed transcripts as untrusted source material and review important summaries before relying on them.