context-audit
Skill19 starsChanged 37 days agoMITView on GitHub
Inspect and reduce the context-window usage of the current session. Use when a session has grown long, behavior degrades, or before starting a new unrelated task in the same session.
19 other projects use a version of this.
Adopted by
Most recently taken up by
- nordz0r/skillstheir own version, first seen Nov 1, 2026
- adamlinscott/claude-skillstheir own version, first seen Oct 30, 2026
- CharlesLadera/claude-custom-skillstheir own version, first seen Oct 25, 2026
- garrytan/gbraintheir own version, first seen Oct 17, 2026
- greglas75/zuvotheir own version, first seen Oct 16, 2026
- edenbuilds/agentlooptheir own version, first seen Oct 13, 2026
Used with
Projects with it also use skillify, refactor and 1 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
MicrosoftSkillskillify
Capture a repeatable process from the current session into a reusable Amplifier SKILL.md skill file. Use when the user wants to create a skill, save a workflow as a skill, turn a process into a reusable skill, or mentions "skillify", "create skill", "make a skill", "save as skill", "capture workflow", "turn this into a skill", "new skill", or wants to automate a repeatable process they just performed.
Also in 3 projects that use this
GitHubSkillrefactor
Surgical code refactoring to improve maintainability without changing behavior. Covers extracting functions, renaming variables, breaking down god functions, improving type safety, eliminating code smells, and applying design patterns. Less drastic than repo-rebuilder; use for gradual improvements.
Also in 3 projects that use this
- MicrosoftSkill
handoff
Compact the current conversation into a handoff document for another agent to pick up.
Also in 4 projects that use this
Other versions19 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 19
- garrytan/gbrain★ 32ksame name
- Neeeophytee/finding-unknowns-skills★ 339same name
- jiahao-shao1/sjh-skills★ 43same name
- Vesely/skills★ 28same name
- WalletConnect/skills★ 14same name
- weellio/gander★ 10same name
- nordz0r/skills★ 8same name
- tstapler/dotfiles★ 8same name
- greglas75/zuvo★ 7same name
- adamlinscott/claude-skills★ 3same name
9 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- broSkill
- closingSkill
- decision-logSkill
- dependency-freshnessSkill
- documentation-updateSkill
- failure-learningSkill
- git-disciplineSkill
- goblin-modeSkill
- project-mapSkill
- run-and-verifySkill
- spec-from-screenshotSkill
- ui-qaSkill
- sdd-updateSkill
- sdd-workflowSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 5 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 19 edited versions we read keep these.
5
Distinctive
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 10
Keep the context window healthy by auditing what is loaded and recommending targeted cleanup, following reference/context-economy.md …
- Distinctive: line 17
Before switching to an unrelated task in the same session.
- Distinctive: line 18
After a noisy investigation (many files read, long tool outputs).
- Very specific: line 38
If compacting, use the harness's compaction command (/compact in Claude Code, Codex and OpenCode) with focus instructions where it accepts …
- Very specific: line 43
Before compacting, write anything durable to its artifact (tasks.json, spec files, decision log, history.md) — artifacts, not chat, are the …
- Distinctive: line 51
No file output by default.
- Distinctive: line 57
Never discard unresolved questions or unrecorded decisions by compacting.
- Distinctive: line 58
Do not write to memory or global configuration.
“Common” is measured: kept by at least 6 in 10 of the 19 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Context audit
Purpose
Keep the context window healthy by auditing what is loaded and recommending
targeted cleanup, following reference/context-economy.md (copied or
linked during onboarding).
When to use
- A session has been running long and responses degrade or lose track.
- Before switching to an unrelated task in the same session.
- After a noisy investigation (many files read, long tool outputs).
- The developer asks "why is context full?" or similar.
When not to use
- Short sessions with no symptoms — auditing costs context too.
- Mid-implementation of an approved task: finish or reach a stable point first, so compaction does not drop in-flight details.
Required inputs
- Current task status (from
tasks.json) and any unresolved questions — these must survive any compaction.
Procedure
- Inspect context usage with the harness's command (
/contextin Claude Code,/statusin Codex, the session view elsewhere — see the table inreference/context-economy.md) and identify the dominant consumers. - Decide: continue, compact, or restart.
- If compacting, use the harness's compaction command (
/compactin Claude Code, Codex and OpenCode) with focus instructions where it accepts them, to preserve: decisions made, current task state, architecture constraints, and unresolved questions. Where it does not, write those to their artifacts first. - Before compacting, write anything durable to its artifact (
tasks.json, spec files, decision log,history.md) — artifacts, not chat, are the source of truth. - Recommend subagents for upcoming noisy exploration instead of doing it in the main conversation.
Output artifact
No file output by default. If durable information was only in chat, the output is updated artifacts (task status, decision entries) written before compaction.
Safety constraints
- Never discard unresolved questions or unrecorded decisions by compacting.
- Do not write to memory or global configuration; this skill only inspects and advises.
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.