maintain-scratch
Skill7 starsChanged 4 days agoMITView on GitHub
Fresh project session, weekly scratch compaction, remove obsolete notes. Triggers on the first project-context load after a fresh session's initial user prompt, once per project per session before substantive work, only when the shared seven-day timer in agents/scratch/timers.json is due; follow-up tasks, rereading instructions, and context-window compaction do not retrigger it.
In this repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- audit-and-resolveSkill
- context-pagingSkill
- coordinate-workSkill
- design-skillSkill
- design-time-triggered-skillSkill
- design-validationSkill
- diagnose-and-repairSkill
- indefinite-loopSkill
- install-dev-skillsSkill
- operate-workflowSkill
- polish-docsSkill
- run-experimentSkill
- contract-disciplineSkill
- documentation-styleSkill
- evidence-disciplineSkill
- git-disciplineSkill
- scratch-spaceSkill
- testing-doctrineSkill
- context-compaction-procedureSkill
- refine-skillSkill
- security-hygiene-checkSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
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 16
Read the index and records.
- Distinctive: line 18
Verify the index, links, and retained context, then stamp only this timer's completion.
≈ “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
Maintain scratch space
Event: first project-context load after the initial user prompt of a fresh agent session. Check once per project per session, before substantive work; follow-up tasks, rereading instructions, and context-window compaction do not retrigger it.
Timer: maintain-scratch in agents/scratch/timers.json, default interval_days: 7. Use the shared timer contract. This trigger is independent of skill self-improvement enablement.
When due:
- Read the index and records. Delete routine change logs, completion receipts, and duplicates of Git history or canonical guidance. Consolidate useful state into current summaries: unresolved work, operational policy, and evidence needed to avoid regression or redundant experiments.
- Archive only evidence with a concrete future use; do not move clutter into an archive by default. Keep active records resumable and repair links. Promote lasting contracts to canonical documentation where appropriate; never discard sole copies of needed evidence or sweep unrelated temporary files.
- Verify the index, links, and retained context, then stamp only this timer's completion. A completed review finding nothing to compact still counts. Include maintenance in the next authorized commit.
Coordinate with active writers. Read-only, interrupted, or unsafe concurrent work leaves the timer unchanged; report the deferral. scratch-space governs ordinary note upkeep between passes, which does not reset this timer.
The host must load the startup instruction. No fresh session means no periodic check; this skill does not install a scheduler.
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.

