apex-raylisted
Install: claude install-skill dobrotacreator/apex-ray
# Apex Ray
## Purpose
Apex Ray is the project's local diff-aware AI review tool. Use it to create deterministic local review reports, run configured LLM review, continue partial coverage, tune repo rules/memory, inspect telemetry, and replay historical PR evals.
## Process
- Run `apex-ray doctor` when setup, config, provider, or analyzer state is uncertain.
- When Apex Ray is configured in a pre-push hook, do not proactively run `apex-ray review` or `apex-ray gate pre-push` as a routine final verification step; let `git push` invoke the hook so the pre-push incremental retry state remains the source of truth.
- For deterministic local review outside pre-push, run `apex-ray review --no-llm` only when the user asks or when diagnosing Apex Ray; default reports are written under `.apex-ray/reports/`.
- When the user asks, the hook is unavailable, or explicit pre-push gate parity is needed before pushing, run `apex-ray gate pre-push`; blocking findings and critical partial coverage are printed to stdout and the full report is written under `.apex-ray/reports/`.
- Do not bypass the configured pre-push gate by default. Use `apex-ray findings suppress` only for confirmed local false positives after checking the finding evidence, current code, and relevant tests or invariants. Provide a concrete objective reason; do not suppress uncertain findings, real defects, or findings merely to get a push through.
- If bypassing is unavoidable, explain why and name the equivalent checks or r