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 yohn1985/jarvisbot --skill youtube-researchgit clone --depth 1 https://github.com/yohn1985/jarvisbotWrote 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/yohn1985/jarvisbot/youtube-research)<a href="https://agentmods.dev/skills/yohn1985/jarvisbot/youtube-research"><img src="https://agentmods.dev/badge/skills/yohn1985/jarvisbot/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/yohn1985/jarvisbot/youtube-research"><img src="https://agentmods.dev/badge/skills/yohn1985/jarvisbot/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 | $0.00025 | $0.00750 |
| Opus 5 | $0.00013 | $0.00375 |
| Sonnet 5 | $0.00005 | $0.00150 |
| Haiku 4.5 | $0.00003 | $0.00075 |
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 9d 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
youtube-research
Ported from the owner's canonical runbook (work/docs/runbooks/youtube-research.md). The
value is the method and judgment, not transcript dumps. skill.py provides the mechanical
tools (search, caption download, VTT cleaning, chunked reading); the mind supplies the synthesis.
Goal
Search the right topic → pick relevant videos → read transcripts → compare patterns across sources → return a clear recommendation + 30-day test plan. A decision, not a book report.
Ground rules (do not relax)
- Read the FULL cleaned transcript for every selected video before concluding. Keyword extraction is orientation only, never a substitute for the full read.
- Treat creator claims as claims, not facts. Separate repeated patterns from one-off claims.
- Don't overquote — summarize in your own words and link sources.
- Note when visual review was not performed (transcripts ≠ watching the frames).
Workflow
- Define the question in one plain sentence + what the final decision must answer.
- Search in clusters (4-8 angle queries) to avoid a single-creator biased sample.
- Build a candidate list (12-20 → pick best 10-15): include competing viewpoints, prefer concrete process over motivation, avoid same-channel duplicates, prefer recent when tactics matter.
- Download transcripts —
skill.py download <video_id>(yt-dlp auto+manual captions → vtt). - Clean captions —
skill.py clean(vtt → de-duplicated plain text). - Read every transcript end-to-end —
skill.py chunk(500-word chunks). Note per video: what they sell, to whom, offer/pricing, funnel, targeting, follow-up, metrics claimed, what's useful, what's inflated, how it applies to us. - Extract evidence by theme (rg/keyword snippets) — cross-check repeated patterns only.
- Score each video — the load-bearing field is "How it applies to us."
- Compare patterns by theme (paid ads / cold email / maps / loom / niche / offer / funnel / follow-up / deliverability): agree? disagree? inflated? directly useful?
- Write the report: Short answer · What I reviewed · Patterns · What I believe is true · What I'd do · 30-day test plan · Metrics to watch · What not to do · Sources.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago First seen · 57 lines · 25 tokens per session scan A 536f0270f702
youtube-research is a skill published in the GitHub repository yohn1985/jarvisbot (1 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 25 tokens to every session and 750 once invoked, about $0.0001 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
agent-ready-features
This skill documents all agent-ready features published by Rise of Machine (https://riseofmachine.com) for automated discovery and integration.
browse-ai-tools
Use the Rise of Machine AI tools directory to discover curated AI tools for founders and SMBs, organised by category.
docs-style-guide
Documentation style guide enforcer by @planetabhi. Applies and reviews the writing style guide when authoring or editing product documentation and tutorials. Use to check prose for voice, tense, word choice, inclusive language, formatting, code block, UI, Markdown, and number/date conventions.
keyboard-accessibility
Load this skill for any project with interactive UI, such as buttons, links, forms, modals, dropdowns, menus, tabs, sliders, toolbars, carousels, or custom widgets, in React or vanilla JavaScript. Never ship an interactive element that cannot be fully operated by keyboard alone. Always ensure a visible focus…
motion-design
Load this skill when adding or reviewing web UI motion such as animations, transitions, entrances and exits, hovers, modals, dropdowns, toasts, page transitions, and micro-interactions. Triggers on "animate", "add a transition", "make it move", "motion", "fade", "slide", "stagger", "open and close". Builds…
ship-ui
Load this skill before shipping a UI change, to statically review the diff for design-craft regressions (accessibility, motion, responsive, and visual or UX drift), correctness and security issues, and leftover debug code, then write a commit message and pull request description that match the repository's…