skill-inventory
Skill340 starsChanged 8 months agoMITView on GitHub
Locates or installs the armory skill-inventory utility, then hands off to its operator guide to measure installed skill descriptions, observed reads and unused managed skills from local OMP, Claude Code and Codex logs. Use when you ask "audit my skill inventory", "how much do skill descriptions cost?", "find unused skills", "which skills were read?", "set up skill-inventory" or "run the skill inventory utility". Verifies the release checksum, asks before downloading and never replaces an existing install. NOT for context-window audits (usage-audit) or catalog installs (skill-library).
1 other project uses a version of this.
Adopted by
Most recently taken up by
- DancinGhost/agent-metaskillstheir own version, first seen Nov 15, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- DancinGhost/agent-metaskills★ 1same name
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 repository94 other files it gives its agents.
- AGENTS.mdAGENTS.md
- adr-writerSkill
- agent-builderSkill
- animateSkill
- api-docs-generatorSkill
- appraiseSkill
- architecture-diagramSkill
- architecture-reviewerSkill
- arxiv-figuresSkill
- arxiv-packageSkill
- arxiv-preflightSkill
- benchmark-runnerSkill
- changelog-composerSkill
- chart-claritySkill
- citation-auditSkill
- codebase-advisorSkill
- code-refinerSkill
- competitive-analyzerSkill
- concept-to-imageSkill
- debug-investigatorSkill
- decision-mapSkill
- dependency-auditSkill
- devils-advocateSkill
- domain-modelingSkill
- engineering-retroSkill
- env-validatorSkill
- estimate-calibratorSkill
- feasibility-assessorSkill
- figure-rhetoricSkill
- figure-table-qualitySkill
- filesystemSkill
- githubSkill
- gpu-optimizerSkill
- handoffSkill
- html-presentationSkill
- html-videoSkill
- humanizeSkill
- idea-validatorSkill
- immuneSkill
- intakeSkill
- kg-builderSkill
- launch-videoSkill
- learning-workspaceSkill
- lightpanda-browserSkill
- linkedin-post-styleSkill
- literature-reviewSkill
- manuscript-provenanceSkill
- manuscript-reviewSkill
- manuscript-typographySkill
- market-analyzerSkill
- marp-slidesSkill
- mcp-to-skillSkill
- md-to-pdfSkill
- migration-risk-analyzerSkill
- milestone-runnerSkill
- module-deepeningSkill
- notebooklmSkill
- package-evaluatorSkill
- package-optimizerSkill
- paper-to-skillSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 37
Read the utility's UTILITY.md and its operator guide (operator-guide.md in the utility's references folder), then follow them.
- Very specific: line 38
Start with the read-only measure --aggregate-only run the guide describes.
≈ “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
Skill inventory
Make the skill-inventory utility available to this agent and hand off to its own documentation. This skill and the utility share a name: the utility is an offline Python tool that ships as an armory utility package, which the Skills CLI and the Claude Code plugin marketplace do not install. This skill is the installable entry point that bridges that gap, without copying the utility's code or measurement rules.
When to use
| Use when | Use instead |
| Auditing installed skill descriptions, observed reads or unused managed skills | — |
| The utility is not installed, or you do not know where it is | — |
| Diagnosing what fills the current context window | usage-audit |
| Searching or installing armory packages | skill-library |
Static conformance report or authored package audit; live effectiveness NOT_MEASURED | package-evaluator |
Triggers group into inventory audits ("audit my skill inventory", "find unused skills", "which skills were read?"), cost questions ("how much do skill descriptions cost?") and setup ("set up skill-inventory", "run the skill inventory utility").
Prerequisites
macOS or Linux with bash, curl, unzip, shasum or sha256sum, Python 3.12+ and uv. Installing needs network access to GitHub once; measurement afterwards runs offline.
Workflow
- Locate an existing install. Run
scripts/ensure-utility.shfrom this skill's directory. It checks~/.claude/utilities/skill-inventory(the armory profile installer's location) and${XDG_DATA_HOME:-~/.local/share}/armory/utilities/skill-inventory. Exit0printsUTILITY_ROOT=andVERSION=; exit3means it is not installed. - On exit
3, ask the owner before installing. Installing downloads and later runs code from the armory rollinglatestrelease. With approval, runscripts/ensure-utility.sh --install, or--install DESTfor another location outside every Git repository; a relativeDESTis resolved against the current directory. - Export the printed root:
export UTILITY_ROOT=.... - Bootstrap the pinned dependency once:
uv run --no-project --script "$UTILITY_ROOT/scripts/skill_inventory.py" --help. - Read the utility's
UTILITY.mdand its operator guide (operator-guide.mdin the utility's references folder), then follow them. They own the commands, private output rules, approval gates and interpretation of results; this skill does not restate them. - Start with the read-only
measure --aggregate-onlyrun the guide describes.recommendwrites only a private proposal.archiveandrestore --applymove files and need the owner's separate, explicit authorization.
Install behavior
ensure-utility.sh --install does the following, and stops with nothing installed if any step fails:
- Downloads
checksums.txtfrom the armorylatestrelease and finds exactly oneskill-inventory-X.Y.Z.utilityentry. - Downloads that archive and compares its SHA-256 with the listed value.
- Unpacks it into a hidden temporary directory beside the destination and checks it contains
UTILITY.mdandscripts/skill_inventory.py. - Renames it to the destination only if nothing exists there, so an interrupted install never leaves a partial copy in place.
The checksum guards against a corrupted or truncated download. It comes from the same release, so it does not prove who published the archive.
The script never upgrades or overwrites an existing install. To upgrade, have the owner compare the installed VERSION with the release, move the old directory aside, then install again.
Worked example
SKILL_DIR="$HOME/.claude/skills/skill-inventory" # where this skill was installed
"$SKILL_DIR/scripts/ensure-utility.sh" # exit 3: not installed
# Owner approves the download.
"$SKILL_DIR/scripts/ensure-utility.sh" --install
# UTILITY_ROOT=$HOME/.local/share/armory/utilities/skill-inventory (expanded)
# VERSION=1.0.0
export UTILITY_ROOT="$HOME/.local/share/armory/utilities/skill-inventory"
uv run --no-project --script "$UTILITY_ROOT/scripts/skill_inventory.py" --help
# Next: follow $UTILITY_ROOT/references/operator-guide.md for the measure run.Output
Report the utility root and version, whether anything was installed, and then the utility's own output as its guide describes. Stdout from the utility is anonymous aggregate JSON; exit 2 means incomplete or refused, not an empty result.
Troubleshooting
| Symptom | Action |
Exit 3 | Not installed. Ask the owner, then rerun with --install. |
could not download ... | Network or GitHub failure; nothing was installed. Rerun later. |
checksum mismatch | Stop. Do not retry blindly or bypass the check; report it to the owner. |
... exists but is not a skill-inventory utility | The destination holds something else. Inspect it with the owner; never delete it to make room. |
uv or Python 3.12 missing | Install uv; uv run provisions the pinned PyYAML on first use. |
Utility exits 2 | Look up each diagnostic code in the utility's measurement.md diagnostic table; do not patch logs or relax checks. |
Files
| File | Purpose |
| scripts/ensure-utility.sh | Locate the utility, or install the released archive after checksum verification |
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.

