agentleFS
Sign inStart free
lorena-bordonaba-paulorena-bordonaba-pau/youtube-agent/.claude/skills/channel-analytics/SKILL.md

channel-analytics

Skill2 starsChanged 48 days agoMITView on GitHub

Audit and diagnose your own channel with real data — performance, retention, traffic, audience, and comparison against earlier snapshots. Triggered by "how's the channel doing?", "audit my videos", "why did it drop", "what's working". In Spanish, "¿cómo va el canal?", "audita mis vídeos", "por qué ha bajado", "qué está funcionando".

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Look at milestones: intro_30s, re_hook_min3, re_hook_min6.

    line 36
  2. Very specific:

    If memory/sop/scripting_sop.md exists, contrast against it.

    line 38
  3. Distinctive:

    The order adapts to the data: If step 1 reveals a sharp drop, go straight to step 4 on the affected video.

    line 48
  4. Very specific:

    If CTR is missing to close the diagnosis, say so and point at tools/ingest_studio_csv.py.

    line 63

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Channel audit

WHEN

Performance diagnosis, periodic report, "why did this video flop?", "what is working and what is not?", comparing periods.

EXECUTION ORDER

  1. Baseline and history python3 tools/yt_report.py --days 28 Gives the overall state, top videos, traffic, audience and geography, and appends a snapshot. If earlier snapshots exist it returns change_since_last_snapshot: that delta is the real diagnosis, not the still photo.
  2. Rank by retention, not only by views python3 tools/yt_top_videos.py --days 90 --by-retention A video with few views and high retention is a format signal; one with many views and low retention is a promise problem.
  3. Where people come from python3 tools/yt_traffic.py --days 28 python3 tools/yt_search_terms.py --days 90 The second is the one that pays: those are the real terms people arrive through. Subscriber dependence above 50% means the channel is not reaching beyond its own base.
  4. Where they leave — on the 3 most relevant videos from step 2 python3 tools/yt_retention.py --video ID Look at milestones: intro_30s, re_hook_min3, re_hook_min6. A drop in the intro is a promise problem; a drop at a re-hook is a structure problem. If memory/sop/scripting_sop.md exists, contrast against it.
  5. Mandatory cross-check before concluding python3 tools/yt_outliers_channels.py --min-ratio 2.0 Without this you cannot tell a channel problem from the whole niche moving.
  6. Compare against what was already known: memory/MEMORY.md, especially any outcome_* files, which hold the prediction made before the video against the real result.

The order adapts to the data. If step 1 reveals a sharp drop, go straight to step 4 on the affected video. The data outranks the manual.

SOURCES

Steps 1-4 youtube_api (aggregates and percentages derived). Step 5 derived. No heuristics: this skill estimates nothing.

OUTPUT

  1. Headline: what changed and since when, with the data's date.
  2. Status table with the delta against the previous snapshot.
  3. Diagnosis: 2-3 causes, each backed by a concrete figure.
  4. Priority action: one only, the highest impact.

If CTR is missing to close the diagnosis, say so and point at tools/ingest_studio_csv.py. See LIMITS.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free