Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add MKirovBG/scribefy-mcp --skill youtube-researchgit clone --depth 1 https://github.com/MKirovBG/scribefy-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mkirovbg/scribefy-mcp/youtube-research)<a href="https://agentmods.dev/skills/mkirovbg/scribefy-mcp/youtube-research"><img src="https://agentmods.dev/badge/skills/mkirovbg/scribefy-mcp/youtube-research/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mkirovbg/scribefy-mcp/youtube-research"><img src="https://agentmods.dev/badge/skills/mkirovbg/scribefy-mcp/youtube-research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00074 | $0.00940 |
| Opus 5 | $0.00037 | $0.00470 |
| Sonnet 5 | $0.00015 | $0.00188 |
| Haiku 4.5 | $0.00007 | $0.00094 |
Grade A, and why
youtube-research scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
YouTube Research with Scribefy
You have four Scribefy tools. Three are free; one costs credits. The core discipline of this skill: spend free calls to make every paid call count.
| Tool | Cost | Use for |
|---|---|---|
search_videos |
Free | Finding candidate videos by query |
get_video_metadata |
Free | Vetting a candidate before extraction |
get_related_videos |
Free | Expanding from a good seed video |
extract_transcript |
1–8 credits by video length — cached extracts are FREE | The actual transcript |
The workflow
1. Search wide (free)
Run search_videos with 2–3 different phrasings of the question — YouTube
search rewards different keyword angles ("PO3 trading strategy" vs "power of
three ICT explained"). Collect ~5–10 candidates.
2. Vet before you spend (free)
Call get_video_metadata on the shortlist. Decide using:
- Duration — extraction cost scales with length (1 credit for short videos up to 8 for very long ones). A focused 12-minute video usually beats a 3-hour podcast that mentions the topic once.
- Caption tracks — the metadata lists every track and marks it
✏ authored or ⚙ auto-generated. Authored captions give cleaner text.
No caption tracks listed → do not attempt extraction (it will fail
with
NO_CAPTIONS); pick another candidate. - Recency and channel — for fast-moving topics, check the upload date before trusting the content.
3. Extract selectively (paid — usually 1–2 videos, not 10)
Call extract_transcript on the best 1–2 candidates first. Read them. Only
extract more if the question is still open.
- Cached transcripts cost 0 credits — popular videos are often already cached, and re-extracting the same video is always free. Never hesitate to re-pull a video you've extracted before.
- For non-English videos, pass the language code of the caption track you saw in metadata. Extract in the original language and translate in-chat — that preserves nuance better than hunting for a translated track.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 86 lines · 74 tokens per session scan A 8e99623c2509
youtube-research is a skill published in the GitHub repository MKirovBG/scribefy-mcp (1 stars, last pushed 3mo ago), licensed MIT. It adds 74 tokens to every session and 940 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
21st-ui
Find, install, and generate UI with 21st.dev. Use when the user asks for a UI component (pricing table, hero, navbar, dashboard, form, etc.), wants design inspiration, needs a brand logo as an SVG component, or wants to generate new UI from a prompt.
webmcp-nexus
A coding guide for adding WebMCP tools to JavaScript or TypeScript projects. WebMCP lets AI agents call selected application functions through a standard interface.
repo-context
Use repo-context-mcp tools to map, search, and pack repository context before large edits or PR review.
social-scrape
Collect public posts and discussion from X, LinkedIn and Hacker News through a local signed-out browser, then report them in chat or write them to a document. Use when the user wants to scrape, monitor or research specific accounts, topics, hashtags or discussions on social platforms.
papi-verify
Use when the user says "verify", "health check", "check papi", "verify cycle", or after completing all cycle tasks before release. Runs a structured health check on the current PAPI cycle state: plan validity, build state, review coverage, branch hygiene, and data consistency.
MCP Integration Assistant
Helps design and implement Model Context Protocol (MCP) server integrations for AI agents.