content-performance-audit — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited content-performance-audit (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Turn a creator's own post history into a ranked, statistically-tested answer to "what actually drives my views?" — instead of guessing, you measure.
This skill pairs with the SocialGPT MCP server (https://mcp.gpt.social/mcp). The MCP provides the data; this skill provides the deterministic analysis and a shareable report. If the SocialGPT tools below aren't available, the user needs to connect the MCP first: https://gpt.social/integrations/mcp
Trigger on requests like: "what drives my views?", "what's working on my account?", "what should I post more of?", "do a performance audit", "why did some videos pop and others flop?"
list_videos isn't available, point the user to the connect page above and stop.
list_videos(limit=50, sort="recent", include_analysis=false)To audit a single connected account, first call list_accounts() and pass its account_id. Save the raw JSON the tool returns to a file named videos.json in the working directory (the whole {"videos": [...]} envelope is fine — the script handles it).
python scripts/analyze.py videos.jsonThe script reads videos.json, runs the tests, prints a Markdown summary to stdout, and writes content-performance-report.html.
offer the user the generated content-performance-report.html to download / open (on Claude.ai it appears as a downloadable file).
For every video with a usable view count (Instagram posts showing 0 plays are treated as missing, since IG hides reel plays), it tests each factor against view count and ranks them by effect size and p-value:
on log-views, with a real two-sided p-value.
standalone) → Kruskal-Wallis rank test, robust to the heavy-tailed view distribution.
It flags factors at p<0.05 (significant) and p<0.01 (highly significant), then translates the top significant factors into plain-English actions. See references/methodology.md for the statistical details and caveats.
rather than invent findings. More posts → more statistical power; 30+ is ideal.
test next*, not guarantees. The report says this explicitly.
grows.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.