Local-first repository intelligence for coding agents: ranked context, likely impact, test routes, risks, and repository benchmarks.
Scoring was not performed