The ForgeLens Brief
20
26
Local-first workflow intelligence for AI coding teams.
Read the risky files before AI edits anything.
ForgeLens scans your repository, maps risky boundaries, and creates compact context reports so coding agents start from auth, data, env, and route hotspots instead of wasting context on random files.
AI coding agents can miss the real risk. They often begin in low-impact files and only later reach auth rules, server actions, data access, and migration edges. That causes wrong assumptions and fragile edits.
ForgeLens changes the starting point. It ranks risky files first, writes focused context maps, tracks baseline snapshots, and flags drift when boundaries move between yesterday and today.
AI Focus Map
Start with admin/API routes, auth boundaries, server actions, and schema files before touching surface UI code.
Baseline + Drift
Save known-good reports and compare with current outputs to catch rule drift around risky edges before next edits.
Safe Env Review
List env key names and config gaps only. Never print secret values in generated reports.
Trust Signals
Release v0.2.0 quality checks you can verify.
CI runs on Linux, macOS, and Windows for each push and pull request.
The packaged npm tarball is built and smoke-tested in CI to catch publish-time CLI issues early.
Run pnpm release-check before publish to validate typecheck, tests, build, and package dry run.
