kit-audit-metrics
Skill6 starsChanged 27 days agoMITView on GitHub
Keep aggregate data work honest — define the question, trace every number to a query, separate fact from interpretation. Use for counts, rates, trends, funnels, reports, or “why did metric X move”.
In this repository20 other files it gives its agents.
- AGENTS.mdAGENTS.md
- rules/AGENTS.mdAGENTS.md
- kit-audit-skillsSkill
- kit-capture-lessonSkill
- kit-challenge-designSkill
- kit-check-blastSkill
- kit-create-prSkill
- kit-debate-decision-treeSkill
- kit-debug-prodSkill
- kit-decisionsSkill
- kit-evolveSkill
- kit-explainSkill
- kit-learn-repoSkill
- kit-minimizeSkill
- kit-optimize-contextSkill
- kit-planSkill
- kit-proveSkill
- kit-reviewSkill
- kit-strip-diffSkill
- kit-triage-commentsSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Not single-session / single-user incident debug (kit-debug-prod)
- Distinctive: line 17
Not inventing charts without query provenance
- Distinctive: line 29
Write a short ANALYSIS DEFINITION block, then proceed.
- Distinctive: line 33
Prefer primary source of truth for that metric
- Distinctive: line 35
Every number in the answer must point to a query/export artifact
- Distinctive: line 41
If an LLM categorizes rows, separate model labels from raw facts
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.
Files
Audit Metrics
Anti-hallucination workflow for aggregate analytics. Source-agnostic (warehouse, logs, Mixpanel, BigQuery, …).
Can it be generic?
Yes. The discipline is generic; connectors differ. Team wiring file may list default sources/exclusions — optional.
What this is NOT
- Not single-session / single-user incident debug (
kit-debug-prod) - Not inventing charts without query provenance
Phase 1 — Define (before querying)
Confirm with the user if missing:
- Question + why it matters
- Environment (default prod — confirm)
- Timeframe (never silently “all time”)
- Exclusions (internal/test users — ask)
- Sampling (full set default; ask before subsetting)
Write a short ANALYSIS DEFINITION block, then proceed.
Phase 2 — Query with provenance
- Prefer primary source of truth for that metric
- Save/reproduce the exact query (or export path)
- Every number in the answer must point to a query/export artifact
Phase 3 — Validate
- Spot-check totals / row counts
- Sanity vs a known baseline if available
- If an LLM categorizes rows, separate model labels from raw facts
Phase 4 — Report
| Fact (with source) | Interpretation (labeled) |
| … | … |
State confidence and what was not checked. No fabricated metrics — if you lack access, say so (kit-prove mindset).
Any external source named in the report (dashboard, doc, dataset) carries a URL and the date checked, or is labeled unverified: <claim> (kit-evidence).
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
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.

