f-catchup
Skill14 starsChanged 20 days agoMITView on GitHub
Bulk-extract Logbook cards from completed jobs that were never captured. Use when card maintenance has fallen behind. Not for a session status refresh (f-status). Not one card from this conversation (f-card). Not close of the current job (f-signoff).
In this repository32 other files it gives its agents.
- AGENTS.mdAGENTS.md
- docs/llms-full.txtllms.txt
- docs/llms.txtllms.txt
- adjudicateSkill
- agent-councilSkill
- converseSkill
- f-archive-searchSkill
- f-archiveSkill
- f-blueprintSkill
- f-boardSkill
- f-cardSkill
- f-complySkill
- f-contextSkill
- f-distillSkill
- f-dreamSkill
- f-greploopSkill
- f-inspectSkill
- f-night-shiftSkill
- f-plancheckSkill
- f-scheduleSkill
- f-searchSkill
- f-setupSkill
- f-shiftSkill
- f-signoffSkill
- f-sketchSkill
- f-statusSkill
- f-surveySkill
- model-councilSkill
- orca-pi-connectSkill
- pack-reviewSkill
- speakSkill
- xreviewSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 3 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
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
- Very specific: line 22
Do not invent learnings. Do not auto-run /f-distill.
- Distinctive: line 23
Do not modify completed jobs.
- Very specific: line 34
After go: root may dispatch bounded code-explorer batches of non-overlapping jobs.
- Distinctive: line 35
Verify each candidate against the source.
- Distinctive: line 36
Write only high-value cards from the installed templates: reusable INSIGHTs, repeatable METHODs, compact BUILDs.
- Very specific: line 37
Update logbook_index.md and logbook_links.md as /f-card Step 9 does: a row per card, counts recounted from the card files (templates …
≈ “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
Owner-invoked, or followed in automatic mode by /f-signoff Step 3.2a. It never fires on description match: a bulk pass writes many card files.
Automatic mode (from /f-signoff Step 3.2a only)
No y/n: the close itself is the go. Only completed jobs that no card names in its job:, at most three per close, oldest first. Every rule below still holds (no invented learnings, verify against the source, no forced BUILD, recount the index and links), except the preview-and-wait in Halt and step 3, and the Report's "Stop": it reports the jobs it looked at, the cards it wrote (every path) and the jobs that got none, to /f-signoff, and returns. Never runs /f-distill. A job whose evidence can't be verified gets no card and is listed for the owner.
Topping Out: Logbook catch-up
Halt
- Preview the gap. Wait for y/n before writing any card.
- Do not invent learnings. Do not auto-run
/f-distill. Distill, if offered later, writes the constitution (the repo profile names it:AGENTS.md, orCLAUDE.mdwhere that is the constitution) only, on the owner's word. - Do not modify completed jobs. Never stage, commit, or push.
Scope
Optional --since YYYY-MM-DD. Otherwise compare completed-job dates (foreman/completed-jobs/*/JOB_*.md) with the latest indexed cards. Read the three templates and logbook_index.md before proposing.
Process
- Enumerate completed jobs in scope and map existing card provenance. Keyword overlap is a hint; coverage is semantic.
- Identify jobs with no coverage or clearly incomplete coverage.
- Present covered vs uncovered. Estimated card counts. Proceed? (y/n)
- After go: root may dispatch bounded
code-explorerbatches of non-overlapping jobs. Children return candidates only and never recursively dispatch. - Verify each candidate against the source. Deduplicate against current IDs.
- Write only high-value cards from the installed templates: reusable INSIGHTs, repeatable METHODs, compact BUILDs. Skip routine jobs (“no cards warranted”). Do not force one BUILD per job.
- Update
logbook_index.mdandlogbook_links.mdas/f-cardStep 9 does: a row per card, counts recounted from the card files (templates excluded), never incremented. Preserve IDs and provenance.
Report
Jobs inspected, cards created/skipped, remaining gaps. Stop. Offer /f-distill as a separate owner step.
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.