skill CX Codex

system-audit · gasserane — Claude Code skill: what it does, how to install

Q: How to install system-audit in Claude Code?

Run `npx skills add gasserane/personal-skills --skill system-audit -a claude-code` in your terminal.

Q: How does system-audit differ from the /test command?

While `/test` runs only the local test harness, system-audit checks across layers. It compares local runs with cloud reports, verifies line budgets, and checks documentation consistency.

Audits multi-agent workspaces for configuration drift, test mismatches, and documentation gaps. Generates a prioritized findings brief without applying automated changes.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓

npx skills add gasserane/personal-skills --skill system-audit -a claude-code

Open on GitHub

What system-audit does

The skill inspects multi-agent workspace layers for silent drift. It executes tests/run_tests.py and matches the results against recent weekly-health-check-*.md logs. Empty directories can cause false passes. To prevent missed failures, the audit diffs check names directly and simulates remote test execution.

The audit examines documentation consistency and agent definitions. Paths are resolved dynamically. The skill queries the test module directly to locate active skills directories, comparing them against the repository clone. On Windows, diff runs with --strip-trailing-cr to suppress newline mismatches. It also flags SKILL_BUDGETS line overruns and overlay files exceeding 35 KB.

The skill does not apply fixes to code or configuration files. It writes a structured action brief listing file locations, line numbers, and suggested fix sequences. You need the underlying Python test harness already configured in your repository. Every change requires manual review.

What it helps with

detect discrepancies between local test runs and cloud routine reports
catch agent skill files approaching line count limits in SKILL_BUDGETS
identify count mismatches across CLAUDE.md files and agent registries
flag agent overlay files exceeding the 35 KB size limit

Example prompts for the agent

/system-audit
Check workspace health and agent consistency with /audit
Run a system check before starting /li curate

How to install system-audit

1Skills CLI (Vercel)

npx skills add gasserane/personal-skills --skill system-audit -a claude-code

2Manually: put the files in place

~/.claude/skills/system-audit/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/system-audit/

1Skills CLI (Vercel)

npx skills add gasserane/personal-skills --skill system-audit -a codex

2Manually: put the files in place

.agents/skills/system-audit/SKILL.md

For all projects use ~/.agents/skills/

1Skills CLI (Vercel)

npx skills add gasserane/personal-skills --skill system-audit -a cursor

2Manually: put the files in place

.cursor/skills/system-audit/SKILL.md

Cursor also reads .agents/skills/ and .claude/skills/

1GitHub CLI

gh skill install gasserane/personal-skills system-audit

2Manually: put the files in place

.github/skills/system-audit/SKILL.md

1Skills CLI (Vercel)

npx skills add gasserane/personal-skills --skill system-audit -a gemini-cli

2Platform CLI

gemini skills install https://github.com/gasserane/personal-skills --path skills/system-audit

Manually: put the files in place

.devin/skills/system-audit/SKILL.md

The legacy .windsurf/skills/ path still works

Skills CLI (Vercel)

npx skills add gasserane/personal-skills --skill system-audit -a opencode

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about system-audit

How to install system-audit in Claude Code?

Run npx skills add gasserane/personal-skills --skill system-audit -a claude-code in your terminal.

Does system-audit apply fixes automatically?

No, the skill never modifies files directly. It generates a prioritized working brief with file paths and line numbers so you can review and apply fixes manually.

How does system-audit differ from the /test command?

While /test runs only the local test harness, system-audit checks across layers. It compares local runs with cloud reports, verifies line budgets, and checks documentation consistency.

system-audit by other authors comes first.

system-audit awrshift Runs a periodic check of the whole agent system against real files and command output: what works, what drifted, what to delete. Built for repos whose rules and memory have piled up. CX Codex +1 ★ 34repo stars setup-matt-pocock-skills Matt Pocock This skill configures a repository for Matt Pocock's engineering skills by setting its issue tracker, triage labels, and domain docs layout. Run it before using the other skills. 1,052,851installs ask-matt Matt Pocock Routes you to the right Matt Pocock skill or flow for the task at hand. Use it when you need a starting point for a feature, prototype, incoming issue, or hard bug. 766,587installs to-tickets Matt Pocock Turns a plan, spec, or conversation into small end-to-end tickets with explicit blockers. Use it to agree on the work breakdown before tickets go to a tracker. 711,230installs writing-plans Jesse Vincent This skill turns requirements for a multi-step task into an implementation plan before code changes begin. It helps when another engineer needs exact files, interfaces, and checks. CX Codex 282,859installs caveman-review JuliusBrussee Turns code review comments into one line per finding: location, problem, fix. Useful when a diff or PR needs a fast pass without the filler. CX Codex +1 411,800installs caveman-compress JuliusBrussee Compresses CLAUDE.md, todo lists, and other prose files into a shorter format while retaining a readable backup. CX Codex +1 411,500installs

From the same set29

vi Multi-agent coordinator for MEL and SRHR research tasks. It builds specialist subagent teams from an approved plan, monitors outputs, and synthesizes the final document. CX Codex +1 652installs toc-builder Builds a Theory of Change with preconditions, testable assumptions and an evidence plan. Useful for programme design and funder submissions. CX Codex +1 645installs instrument-designer Builds the data-collection tool itself: questionnaires, interview and focus-group guides, observation checklists. Useful once indicators exist but nothing measures them. CX Codex +1 645installs wrap-up This skill closes a working session: it gathers git state, errors and open tasks, delivers a status report, then commits and pushes pending work. CX Codex +1 639installs grade-system This skill grades an MEL/SRHR system across components, quality dimensions, and adapted OECD-DAC criteria. Use it for a structured review of risks and evidence. CX Codex +1 633installs learning-agenda Builds IPPF learning agenda documents: annual federation agendas, quarterly RAG status dashboards, and adaptation templates for member associations under USAID and Patton rules. CX Codex +1 628installs learning-product Turns a Researcher Evidence Brief into one of four outputs: a two-page MA brief, a workshop pack, a regional brief or a donor memo. Citations stay limited to the brief. CX Codex +1 626installs localise Adapts finished English deliverables into Romanian, French, Spanish, or Russian with glossary checks, back-translation QA, and deck reviews. CX Codex +1 620installs