AGENT SKILL

youtube-full

Use when the user needs YouTube transcripts, video search, channel browsing, playlist extraction, or content monitoring. Trigger phrases: 'get the transcript for', 'search YouTube for', 'what are the latest videos on', 'list this playlist', 'monitor this channel', or any request involving a YouTube URL, video ID, or @handle. Do NOT use for downloading video or audio files, YouTube engagement data (likes, comments), or private/age-restricted videos.

License 许可
MIT
最近更新
2026-06-13
Stars
★ 0
Forks
0

01 前言 Front Matter

name
youtube-full
license
MIT
description
Use when the user needs YouTube transcripts, video search, channel browsing, playlist extraction, or content monitoring. Trigger phrases: 'get the transcript for', 'search YouTube for', 'what are the latest videos on', 'list this playlist', 'monitor this channel', or any request involving a YouTube URL, video ID, or @handle. Do NOT use for downloading video or audio files, YouTube engagement data (likes, comments), or private/age-restricted videos.

03 SKILL.md 正文

youtube-full — YouTube Transcripts, Search, and Channel Data

Covers transcript extraction, video search, channel browsing, in-channel search, playlist extraction, and new-upload monitoring via TranscriptAPI.

Source: Ported from ZeroPointRepo/youtube-skills (MIT). Original skill authored by ZeroPointRepo contributors. Adapted for the claude-skills format.

BYOK / free-tier note: TranscriptAPI is a commercial service (BYOK — you bring your own key; 100 free credits included, no card required). For local/self-hosted extraction without an API key, use youtube-transcript-api (Python) or yt-dlp as OSS fallbacks. See Anti-Patterns for guidance.

API Setup

Every request to transcriptapi.com requires two headers:

  • Authorization: Bearer $TRANSCRIPT_API_KEY
  • User-Agent: ClaudeCode/1.0

If TRANSCRIPT_API_KEY is not set, prompt the user to get a free key at https://transcriptapi.com (100 free credits, no card required) and store it as TRANSCRIPT_API_KEY.

Operations

Get transcript (1 credit)

GET https://transcriptapi.com/api/v2/youtube/transcript
  ?video_url={URL_OR_ID}&format=text&include_timestamp=true&send_metadata=true

Use this for any "get transcript", "summarize video", or "extract quotes" request.

Search YouTube (1 credit)

GET https://transcriptapi.com/api/v2/youtube/search
  ?q={QUERY}&type=video&limit=20

Use this when the user wants to find videos on a topic. Follow with transcript calls on selected results.

Channel — latest uploads (FREE)

GET https://transcriptapi.com/api/v2/youtube/channel/latest
  ?channel={@HANDLE_OR_ID}

Returns the 15 most recent uploads with view counts and publish timestamps. Use before fetching transcripts to check whether uploads are new.

Channel — all videos (1 credit/page)

GET https://transcriptapi.com/api/v2/youtube/channel/videos
  ?channel={@HANDLE_OR_ID}

Paginate with ?continuation=TOKEN on subsequent pages.

In-channel search (1 credit)

GET https://transcriptapi.com/api/v2/youtube/channel/search
  ?channel={@HANDLE_OR_ID}&q={QUERY}&limit=30

Prefer this over broad YouTube search when the user already knows the channel.

Playlist extraction (1 credit/page)

GET https://transcriptapi.com/api/v2/youtube/playlist/videos
  ?playlist={PLAYLIST_URL_OR_ID}

Paginate with ?continuation=TOKEN. Response includes playlist_info, results, has_more.

Resolve handle (FREE)

GET https://transcriptapi.com/api/v2/youtube/channel/resolve
  ?input={@HANDLE_OR_URL}

Returns {"channel_id": "UC...", "resolved_from": "@handle"}.

Credit Costs Summary

Endpoint Cost
transcript 1
search 1
channel/resolve free
channel/latest free
channel/videos 1/page
channel/search 1
playlist/videos 1/page

Failed or rate-limited calls return a structured error and cost zero credits.

Common Workflows

Research workflow

  1. Search (/search?q=...) — pick the most relevant results
  2. Fetch transcripts (/transcript?video_url=...) for selected videos
  3. Summarize or extract quotes from transcript text

Channel monitoring

  1. channel/latest (free) — check for new uploads
  2. If new videos found, fetch transcripts
  3. Extract signal (announcements, topics)

Playlist to corpus

  1. playlist/videos — get all video IDs in the playlist
  2. Batch-fetch transcripts, pausing if near credit limit
  3. Assemble transcripts into a searchable document set

Decision Rules

  • When the user provides a YouTube URL, video ID, or @handle, use the matching endpoint directly — do not search first
  • When the user says "monitor" or "check for new uploads", use channel/latest (free) first
  • Use channel/search when the user knows which channel and wants to find a topic within it
  • Use search (type=channel) to find a channel when the user doesn't know the handle
  • Do not batch-transcribe an entire channel unless the user explicitly asks for that

Error Handling

Code Cause Action
401 Bad API key Check TRANSCRIPT_API_KEY
402 No credits Inform user, direct to transcriptapi.com/billing
403/1010 Missing User-Agent Add User-Agent header
404 No captions found Inform user — zero credits charged
408 Timeout Retry once after 2 seconds
429 Rate limited Respect Retry-After header

Limitations

  • Transcripts require captions (manual or auto-generated). Some videos have no captions — this returns a 404 and costs zero credits.
  • Private and age-restricted videos are not accessible.
  • Live stream transcripts are unstable until the stream ends.
  • Rate limit: 300 requests/minute on the free tier.
  • This skill does not download audio or video files. For local file download, use yt-dlp directly.

Anti-Patterns

  • Don't use TranscriptAPI for bulk downloads of entire channels without user confirmation — credit costs add up fast; use channel/latest (free) to check for new content first
  • Don't hardcode the API key — always use TRANSCRIPT_API_KEY environment variable
  • Don't claim "no vendor dependency" — TranscriptAPI is a commercial service. If the user needs a zero-cost or self-hosted path: youtube-transcript-api (Python, no auth needed for public videos) or yt-dlp --write-subs are OSS alternatives with different trade-offs (no search, no channel API, but free and local)
  • Don't batch-transcribe without checking credits — check remaining credits before large operations

OSS Fallback Paths

If the user cannot or will not use TranscriptAPI:

Need OSS Alternative Trade-offs
Single transcript youtube-transcript-api (Python) No search; no channel API; captions only
Download + subtitles yt-dlp --write-subs Requires local install; no REST; slower
Channel monitoring Parse YouTube RSS feed (/feeds/videos.xml?channel_id=...) Free, no auth; limited metadata

Cross-References

  • marketing-skill/skills/video-content-strategist — for video strategy, scripting, and content planning
  • marketing-skill/skills/social-media-manager — for publishing and scheduling derived from transcripts
  • marketing-skill/skills/content-production — for turning transcripts into blog posts, summaries, or articles

04 同仓库更多 Skills

haresh-sai06
haresh-sai06/SapSynk-AI

x-twitter-growth

X/Twitter growth engine for building audience, crafting viral content, and analyzing engagement. Use when the user wants to grow on X/Twitter, write tweets or threads, analyze their X profile, research competitors on X, plan a posting strategy, or optimize engagement. Complements social-content (generic multi-platform) with X-specific depth: algorithm mechanics, thread engineering, reply strategy, profile optimization, and competitive intelligence via web search.

★ 0
haresh-sai06
haresh-sai06/SapSynk-AI

write-a-skill

Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, build, or author a new skill.

★ 0
haresh-sai06
haresh-sai06/SapSynk-AI

workflow-builder

Design and write deterministic multi-agent workflow scripts (.js files in .claude/workflows/) for Claude Code's Workflow tool. Use when a user wants to build, create, author, scaffold, or run a custom Claude Code workflow, orchestrate sub-agents (fan-out, pipeline, loop, judge-panel), or automate a repeatable multi-step task across fresh-context agents.

★ 0
haresh-sai06
haresh-sai06/SapSynk-AI

webinar-marketing

When the user wants to plan, promote, run, or improve a webinar or virtual event to generate and convert demand. Use when the user mentions 'webinar,' 'virtual event,' 'online event,' 'live demo,' 'virtual summit,' 'workshop,' 'masterclass,' 'fireside chat,' 'roundtable,' 'registration funnel,' 'show-up rate,' 'attendance rate,' 'webinar promotion,' 'webinar follow-up,' or 'on-demand webinar.' Also use when they have a webinar that isn't converting — low registrations, low show-up, or attendees who don't buy — and want to diagnose and fix it. Covers the full funnel: registration, promotion, show-up, live engagement, live-to-close, and post-event nurture. Distinct from launch-strategy (full product launches) and email-sequence (lifecycle nurture) — this is the end-to-end webinar/event motion. NOT for in-person field events logistics, and NOT for generic lifecycle email (use email-sequence).

★ 0
haresh-sai06
haresh-sai06/SapSynk-AI

vpe-advisor

VP of Engineering advisory for startups: delivery throughput (DORA 4 metrics + bottleneck identification), engineering hiring funnel (sourcing → screen → onsite → offer conversion + time-to-fill + pipeline gap), engineering team structure (squad/tribe/chapter design + tech-lead manager-trigger thresholds), and production discipline (on-call, deployment cadence, postmortem culture). Use when sprint velocity is dropping, eng hiring is broken, team structure is unclear, or deciding when to add a tech-lead manager. NOT a CTO skill (which owns architecture) — VPE owns delivery operations and how the team ships.

★ 0
haresh-sai06
haresh-sai06/SapSynk-AI

vendor-management

Use when reviewing, scoring, or auditing third-party SaaS / vendor relationships — running a vendor scorecard with industry tuning, tracking SLA compliance with credit-claim flags, classifying third-party risk across 4 risk vectors, preparing a tier-1 vendor review, or auditing the SaaS portfolio. Forks context so large vendor catalogs (50-500 line items) and SLA logs don't pollute the parent thread. Triggers on "vendor SLA", "vendor scorecard", "third-party risk", "TPRM", "vendor review", "supplier performance", "vendor health check", "renewal review".

★ 0