DemoSimulated data. Deploy your own DashClaw instance to connect real agents.

Branch Finish

Experimental

The MoltFire + Claude Code branch-finish loop — render the finish prompt, check coding standards, verify capability health, and dry-run the quality gate before you merge.

Live

Finish templates

Standards

Unhealthy capabilities

Stale checks

Branch-finish templates

Standards knowledge

Capability HealthOpen

Loading capability posture…

Branch-finish quality (dry-run)

Dry-run a contains scorer against a sample branch-finish outcome. Side-effect-free — no eval row is written.

Required keywords (comma-separated)Sample outcome

Recent renders

Recent decisions