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 not0lucky/tubescout --skill yt-idea-minegit clone --depth 1 https://github.com/not0lucky/tubescoutWrote 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/not0lucky/tubescout/yt-idea-mine)<a href="https://agentmods.dev/skills/not0lucky/tubescout/yt-idea-mine"><img src="https://agentmods.dev/badge/skills/not0lucky/tubescout/yt-idea-mine/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/not0lucky/tubescout/yt-idea-mine"><img src="https://agentmods.dev/badge/skills/not0lucky/tubescout/yt-idea-mine.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 | $0.00084 | $0.00751 |
| Opus 5 | $0.00042 | $0.00376 |
| Sonnet 5 | $0.00017 | $0.00150 |
| Haiku 4.5 | $0.00008 | $0.00075 |
Grade A, and why
yt-idea-mine 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 12d 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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
yt-idea-mine
YouTube is where founders show receipts (dashboards, revenue, playbooks) that never appear in Reddit threads. Mine it systematically instead of watching it.
Process
- Map demand.
get_search_suggestionson the seed niche term, then on the 3–5 most interesting suggestions (suggestions-of-suggestions). Autocomplete = what people actually type; note language variants (Arabic/French/Spanish suggestions = underserved non-English demand). - Find the case studies.
search_videoswith 2–4 query shapes:"<niche>" how I make/"<niche>" $ per month(revenue case studies)"<niche>" tutorialsorted byview_count(education demand)- upload window
monthoryear, sorted byview_count(what's rising)
- Read the best evidence. Pick the 3–6 most-viewed case-study videos (skip pure
hype: no numbers in title/snippet, or clickbait engagement patterns) and pull
get_transcripts. Extract: business model, revenue claimed, method, tools, and — most valuable — complaints and gaps the creator mentions in passing ("the annoying part is…", "there's no good tool for…"). - Synthesize ideas. Each idea must cite its evidence: the demand signal (suggestions/views) + the pain source (video + what was said). Ideas without both get cut. Rate each: demand evidence / competition seen / effort to test.
- Filter through the user. Rank ideas against what the conversation reveals about the user's actual capabilities and assets (skills, infrastructure, audience, domain knowledge) — an idea that's a 6/10 in general but sits on the user's unfair advantage outranks a generic 8/10. Say when that reranking happens and why.
- Report. Ranked ideas with receipts, then the discarded ones with the reason (saturation, no demand signal, single-source, bad founder-fit).
Rules
- Cite only URLs returned by tubescout tools in this conversation — never write a YouTube URL or video ID from memory.
- Never launder a creator's recycled idea as evidence — a video saying "build X, it's a great idea" is not a signal; a creator complaining about a missing tool is.
- Note each source's incentive (most idea-listicle channels sell idea databases).
- 10–15 tool calls is the normal budget; go deeper only if the user asks.
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.
- 12d ago First seen · 48 lines · 84 tokens per session scan A d240fa32e67a
yt-idea-mine is a skill published in the GitHub repository not0lucky/tubescout (0 stars, last pushed 16d ago), licensed MIT. It adds 84 tokens to every session and 751 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
youtube-research
Research topics across YouTube videos efficiently using Scribefy's MCP tools — search for candidates, vet them with free metadata before spending credits, extract transcripts selectively, and synthesize timestamped answers. Use when researching a topic on YouTube, summarizing or comparing videos, pulling quotes or…
opik-diagnose
Surface the Opik traces worth a developer's attention, ranked by signal — Diagnostics issues first, then errors, failed tool calls, latency, regressions, and low online-eval scores. With the Opik MCP connected it lists the project's agentinsightsissue entities, offers to turn Diagnostics on when the project has it…
client-scripts
Write ServiceNow client scripts (onLoad/onChange/onSubmit/onCellEdit) using gform, guser, GlideAjax, field visibility/mandatory toggles, and validation with debounced server calls.
split-to-prs
Split current work into small reviewable PRs. Use when the user asks to split a chat, set of changes, branch, or PR.
agreement-setup
Set up a bKash tokenized agreement for repeat charges and check its status.
planfix-api
A Planfix CRM integration for managing tasks and contacts. A CRM is a system for organizing customer and business relationships.