Run six-dimension LLM reviews and automated web app health checks on codebases.
npx skills add https://github.com/vsokh/maestro --skill codehealth-vsokh
Please help me install this Agent Skill. Skill: codehealth Source: https://github.com/vsokh/maestro/tree/main/.claude/skills/codehealth Command: npx skills add https://github.com/vsokh/maestro --skill codehealth-vsokh
Codehealth identifies how verifiable a codebase is so that AI agents can safely change code, run validation gates, and know whether their changes are correct. It surfaces gaps in type safety, tests, error handling, security, architecture, and domain logic, and separates an automated health check from the LLM-driven quality assessment. The Skill produces actionable findings and a prioritized backlog so teams and orchestrators can fix agent-readiness problems.
Invoke the codehealth skill with /codehealth to run a full audit that scores six dimensions, performs a web app health check, and generates a prioritized Quality backlog for dev-manager.
High-intent search queries and answers about installing and using this skill.
Assess code readiness for autonomous AI agents by running a six-dimension LLM review and automated health check to score type safety, tests, error handling, security, architecture, and domain logic. This produces a prioritized backlog of Quality tasks to fix verification gaps before enabling agents.
An agent-readiness check measures whether a codebase is verifiable enough for an AI agent to safely change code and run validation gates. You need it before enabling autonomous change agents to ensure tests and validation gates will catch regressions.
Run an automated web app health check by executing builds, linters, tests, coverage analysis, bundle analysis, dev-server checks, Lighthouse, and dependency audits. This captures objective metrics and combines them with an LLM-driven quality assessment to produce actionable findings.
You can run an incremental code quality review on recent changes using diff mode. Diff mode performs an incremental re-scan of the codebase rather than a full audit, allowing you to quickly re-evaluate agent readiness after specific modifications.
The codehealth skill integrates with orchestrator workflows by writing latest scores, historical trends, and backlog artifacts. It creates Quality tasks in dev-manager state, allowing the orchestrator to directly act on agent-readiness remediation items.
LLM-driven code quality reviews focus on six contextual dimensions but do not replace objective automated health checks. To get a complete assessment of agent readiness, you must combine the LLM review with automated builds, tests, and dependency audits to capture both subjective and objective metrics.
securityreviewaiAnalyzes Git repositories to produce summaries of their structure, risks, and key components.
jmaglyAnalyze lcov, cobertura, and istanbul coverage reports to identify gaps and recommend prioritized tests.
IMBurbankCoordinate Cloud Run deployments across dev, staging, and production environments.
reomgitTest web applications and generate detailed QA reports with repro steps.
