Java audit
A Claude Code plugin and marketplace providing a skills library for high-quality Java development.
npx -y skills add mtkhawaja/java-skills --skill java-auditAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use when asked to audit, sweep, or review Java code in bulk against the house conventions — the whole repository, the current branch's diff, or named paths/classes. Also invoked directly as /java-skills:java-audit [paths | diff]. Read-only; reports findings without modifying files.
SKILL.md
1.9 KB, as published. Nobody here has run it
Java Audit
Audit Java code in the current repository against the java-skills conventions.
Scope
Determine the scope from the arguments: $ARGUMENTS
- Empty: audit the repository's main Java source — every module's production code and tests. If that exceeds ~40 files, audit a representative slice per module (entry points, core services, persistence, tests, build files) and say what was sampled.
diff: audit only files changed on the current branch (staged + unstaged + commits not on the default branch).- Paths: audit exactly the named files/directories, plus their immediate tests.
How to run it
Dispatch the java-reviewer agent from this plugin with the scope above; it loads the relevant
skills and produces severity-grouped findings. For very large scopes, dispatch one agent per
module in parallel rather than one giant sweep.
If the agent is unavailable, do it inline: read the relevant SKILL.md files from this plugin's
skills/ directory (java-development always; java-testing for tests; java-concurrency,
java-observability, java-maven, java-gitignore as the file types warrant), then review every
in-scope file against the skills' rules and "Red flags — stop" lists.
Report
Whether delegated or inline, end with a single consolidated report:
- One-line verdict (clean / nits only / needs changes).
- Findings grouped by severity (blocker / should fix / nit), each with
file:line, the rule and skill violated, and a concrete fix. - What was checked and found clean, and what was out of scope.
Do not modify any files — this is a read-only audit.