keel-stats
Skill9 starsChanged 51 days agoMITView on GitHub
Render the ritual telemetry visually — turns .claude/ritual-log into reports/ritual-stats.md with PLAN.md-style colored Mermaid interval boxes (session/compact boundaries) + a counts table, plus git-derived PROCESS metrics (process-only commit share · LOC by class · always-loaded KB · last test-log age · Review depth). Answers "which skills/commands/hooks ran, how often" and "how much of this project is process".
In this repository17 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- keel-adoptSkill
- keel-agent-team-createSkill
- keel-agent-team-startSkill
- keel-auditSkill
- keel-autopilotSkill
- keel-compactSkill
- keel-continueSkill
- keel-distillSkill
- keel-handoverSkill
- keel-phase-reviewSkill
- keel-pilotSkill
- keel-planSkill
- keel-researchSkill
- keel-teamSkill
- keel-tidySkill
- keel-updateSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 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.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 13
Run the deterministic generator — never parse the log by hand:
≈ “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.
Files
/keel-stats — see what ran, when, how often
The ritual-log hook records every Skill-tool call, user-typed command (custom = command …
via UserPromptExpansion; built-ins like /compact = typed /… via UserPromptSubmit — first
token only, user prose never logged), compact boundary, session start and hook BLOCK into
.claude/ritual-log (machine-local, git-ignored, self-trimmed). This skill renders it for humans.
- Run the deterministic generator — never parse the log by hand:
python3 .claude/ritual-report.py - Open / point the user at
reports/ritual-stats.md— the Mermaid timeline renders on GitHub and in the VSCode markdown preview: one colored box per interval (green = session start, amber = manual compact, red = auto compact; PLAN.md's palette), event counts inside, full detail in the table below it. - Quote the headline numbers (totals line) back to the user in one sentence — and the
Process metrics block under it: share of commits touching no product/test code · tracked
lines by class (product / tests / process) · always-loaded KB per session · age of the last
make testlog ·## Reviewdepth and oldest date. These are derived from git + the tree by the same script (never typed by hand); a project that never counted them tends to be surprised (field: 70% process-only commits, reports 3.8x the product, ~234 KB per session). - Scope honesty: the log is machine-local and trimmed to ~1000 lines — long-horizon history lives in git/HANDOVER, not here. Do not commit the report unless the user asks (it is a regenerable artifact; the compact-gate ignores it either way).
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.