ProjectDXAI/labrat

Autonomous multi-branch research lab. Branches compete for compute budget. The system converges on what works.

238Stars on the repository
9Mods indexed here, across every type
24d agoLast push, which is what freshness is scored on
MITLicence, which decides whether bodies are shown

audit-candidate

01

ProjectDXAI/labrat

Command Claude Code

Walk the highest-signal suspicious candidate through an implementation audit.

238 24d ago A 0 tokens original MIT

consolidate

02

ProjectDXAI/labrat

Command Claude Code

Write a compact checkpoint summary of the current frontier.

238 24d ago A 0 tokens original MIT

frame-break

03

ProjectDXAI/labrat

Command Claude Code

Trigger a frame-break cycle only after cheap probes and audits are no longer productive.

238 24d ago A 0 tokens original MIT

next

04

ProjectDXAI/labrat

Command Claude Code

Run python scripts/operatorhelper.py next-prompt --runner claude --phase auto from this lab's root directory, read the prompt it prints, and then execute that prompt's instructions.

238 24d ago A 0 tokens original MIT

synthesize

05

ProjectDXAI/labrat

Command Claude Code

Synthesize what the recent evaluations mean before dispatching more work. This step closes the gap between "config sweeper" and "research agent.".

238 24d ago A 0 tokens original MIT

why-stuck

06

ProjectDXAI/labrat

Command Claude Code

Diagnose why the frontier is not moving.

238 24d ago A 0 tokens original MIT