Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/AlekseiUL/hermes-researcher-agentnpx agentmods add skills/alekseiul/hermes-researcher-agent/youtube-research-packWrote 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/alekseiul/hermes-researcher-agent/youtube-research-pack)<a href="https://agentmods.dev/skills/alekseiul/hermes-researcher-agent/youtube-research-pack"><img src="https://agentmods.dev/badge/skills/alekseiul/hermes-researcher-agent/youtube-research-pack/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/alekseiul/hermes-researcher-agent/youtube-research-pack"><img src="https://agentmods.dev/badge/skills/alekseiul/hermes-researcher-agent/youtube-research-pack.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00044 | $0.01588 |
| Opus 5 | $0.00022 | $0.00794 |
| Sonnet 5 | $0.00009 | $0.00318 |
| Haiku 4.5 | $0.00004 | $0.00159 |
Grade A, and why
youtube-research-pack 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 7d 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 — 169 lines — stays where its author put it; the contents beside it link to each section on GitHub.
YouTube Research Pack
Purpose
Research public YouTube evidence without turning access failures into invented facts. Use the lightweight repository helper for quick work and the optional youtube-intelligence-stack companion for persistent multi-layer runs.
This pack is public-only by default. It never requires cookies, login, OAuth, API keys, browser-profile extraction, private analytics, or account actions.
Choose the smallest lane
Quick search
Use for finding a bounded set of public videos:
python3 tools/youtube_research.py search "<public query>" --limit 5 --json
The helper rejects queries containing email addresses, local paths, phone-like values, and common credential shapes before network access.
Video verification
Use for current public metadata and subtitle availability:
python3 tools/youtube_research.py video "<public video URL or ID>" --json
Treat views, likes, and comments as volatile observations. Record the collection date. Creator-side CTR, retention, traffic source, subscriber conversion, and revenue are unavailable unless the creator supplies them separately.
Channel or playlist scan
python3 tools/youtube_research.py channel "https://www.youtube.com/@handle" --tab videos --limit 10 --json
python3 tools/youtube_research.py channel "https://www.youtube.com/@handle" --tab shorts --limit 10 --json
python3 tools/youtube_research.py playlist "<public playlist URL>" --limit 10 --json
The channel URL is normalized to an explicit videos, shorts, or live tab. Flat listings are discovery evidence only; inspect shortlisted videos separately for current metrics.
Transcript
python3 tools/youtube_research.py transcript "<public video URL or ID>" --languages "en.*,en,ru.*,ru" --json
The helper writes subtitle files only inside a temporary directory, converts one public VTT track to bounded text, and removes the temporary directory. Automatic captions may be wrong. Transcript absence means no_subtitles, not that the video lacks relevant content.
What ships with it
1 file 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.
- 7d ago First seen · 169 lines · 44 tokens per session scan A 25598b252aa6
youtube-research-pack is a skill published in the GitHub repository AlekseiUL/hermes-researcher-agent (57 stars, last pushed 7d ago), licensed MIT. It adds 44 tokens to every session and 1,588 once invoked, about $0.0002 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 2026-09-05.
Other skills, from other repositories
humanize-automation
Human-like mouse, keyboard and scroll behavior for behavioral bot bypass.
hermes-tweet
Use Xquik in Hermes Agent for public X research, monitoring, thread summaries, creator discovery & approved actions. Not affiliated with X Corp. Use when the user requests X data or a named action. Trigger with "search X", "monitor X", "post tweet", or "X trends".
browser-harness-authoring
Use when mapping a repeatable website workflow into a verified Hermes skill so later runs can follow known steps instead of rediscovering the site. Surveys browser compatibility, semantic targets, failure modes, recovery paths, decision gates, and expiry using dummy data and no irreversible submissions.
obsidian-memory-architecture
Use when designing, setting up, or maintaining an Obsidian vault as Hermes Agent's durable knowledge layer. Routes facts, conversation history, documents, procedures, and daily logs to the correct Hermes or vault system without duplicating everything into the prompt.
hermes-diagnostic-review
Use when running a read-only diagnostic review of recent Hermes sessions to find recurring mistakes, failed tool calls, and repeated fixes, then propose suggestion-only improvements and reusable skills. Human-gated; never auto-applies.
holix-cron
Schedule recurring agent tasks via Holix built-in gateway cron (not crontab or custom scripts).