youtube-transcript
Extract transcripts from YouTube videos. Use when the user asks for a transcript, subtitles, or captions of a YouTube video and provides a YouTube URL (youtube.com/watch?v=, youtu.be/, or similar). Supports output with or without timestamps.
90
86%
Does it follow best practices?
Impact
96%
2.13xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that clearly communicates what the skill does, when to use it, and includes strong natural trigger terms. It covers multiple synonyms (transcript, subtitles, captions), specifies URL format variations, and mentions the timestamps feature. The description is concise yet comprehensive with no unnecessary fluff.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists specific concrete actions: 'Extract transcripts from YouTube videos' with details about timestamp support and multiple URL format support. | 3 / 3 |
Completeness | Clearly answers both what ('Extract transcripts from YouTube videos, supports output with or without timestamps') and when ('Use when the user asks for a transcript, subtitles, or captions of a YouTube video and provides a YouTube URL'). | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'transcript', 'subtitles', 'captions', 'YouTube video', 'YouTube URL', plus specific URL patterns like 'youtube.com/watch?v=', 'youtu.be/'. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche: YouTube transcript extraction. The specific URL patterns and YouTube-specific terminology make it very unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
72%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A solid, focused skill that provides clear executable commands and specific output handling instructions. The main weaknesses are minor verbosity in the Notes/URL sections and a lack of error handling guidance (e.g., what to do when captions are unavailable). The 'CRITICAL: YOU MUST NEVER MODIFY' instruction is oddly juxtaposed with the next bullet that says to clean up formatting, which could cause confusion.
Suggestions
Clarify the apparent contradiction between 'NEVER MODIFY THE RETURNED TRANSCRIPT' and 'you SHOULD clean it up so that it is arranged by complete paragraphs' — specify that reformatting whitespace/line breaks is allowed but content words must not be changed.
Add brief error handling guidance: what should Claude do or tell the user when captions are unavailable or the script fails?
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient but the 'Notes' section contains some unnecessary information (e.g., explaining caption fallback behavior is something Claude can infer). The 'Supported URL Formats' section is borderline unnecessary since the script presumably handles parsing internally. | 2 / 3 |
Actionability | Provides concrete, copy-paste ready bash commands for both default and timestamped usage. The output handling instructions are specific and actionable (file naming convention, paragraph formatting rules). | 3 / 3 |
Workflow Clarity | The workflow is relatively simple (run script, handle output), but the output section mixes critical behavioral constraints ('NEVER MODIFY') with formatting instructions and file-saving logic without clear sequencing. There's no error handling guidance for when captions aren't available. | 2 / 3 |
Progressive Disclosure | For a simple, single-purpose skill under 50 lines, the content is well-organized into logical sections (Usage, Defaults, Output, Notes) with clear headers. No need for external file references given the scope. | 3 / 3 |
Total | 10 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- intellectronica/agent-skills
- Commit
9b0e00a
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.