Commit History

  1. final mvp created

    98b952a
  2. feat(phase10): ABScriptEnv, ContrastiveReward, A/B rollout fn, 25 tests PASS, gate PASS

    cfe83fc

  1. Phase 8 implemented

    dfa9070
  2. Phase 7 implemented

    0e4f105
  3. Phase 6 implemented

    5c28dc0
  4. Phase 5 implemented

    e6b6793
  5. feat(phase4): critic escalation engine, difficulty tracker, env wiring, gate PASS

    ebae6ab
  6. phase2 implemented

    258783b
  7. feat(phase1): OpenEnv scaffold + R1/R2 rewards — PHASE 1 GATE PASS⁠

    vajeeda +1: Claude Sonnet 4.6
    41ea373
  8. feat(phase0): add Groq backend, set as default (llama-3.3-70b-versatile)

    5fcd0ee
  9. feat(phase0): switch default model to Qwen2.5-1.5B-Instruct

    bb30917
  10. refactor(phase0): replace Anthropic SDK with LLMBackend abstraction (default Qwen)

    639b641
  11. feat(phase0): add run_critic_gate CLI with rich output and fixture saving

    7414383
  12. feat(phase0): scaffold directories and requirements

    806216c