docs(prd): cross-family verification rule + STATUS truth-pass (erp#56/#63/#39) #31

Merged
arcodange merged 1 commits from arcodange/harness-portability-qa into main 2026-07-18 20:23:14 +02:00
Owner

Paired with erp#69 (closed erp#63) — the harness-portability spike's PRD half.

  • qa-strategy › Independent verification: codifies the cross-family rule — with Mistral (vibe -p, mistral-medium-3.5) and Ornith 35B (hermes MLX) admitted to verifier duty by verdict parity with the Claude baseline on both canonical verifier tests (erp#63 evidence, blind-judged 16/16 unanimous), the independent verifier SHOULD be a different model family than the builder.
  • STATUS truth-pass: #63 2026-07-18 (PR erp#69, harness home erp:fleet/harness/), #56 (PR erp#68 — authored end-to-end by the Mistral builder bench, 0 code corrections), #39 golden set built on local branch arcodange/golden-set (push+PR = operator step — blocked by the session permission gate on real financial documents), PR-log rows, P3 → 🟡 in progress.

🤖 Generated with Claude Code

https://claude.ai/code/session_01VRShc4QhLLU73FLHx9vskh

Paired with [erp#69](https://gitea.arcodange.lab/arcodange-org/erp/pulls/69) (closed erp#63) — the harness-portability spike's PRD half. - **qa-strategy › Independent verification**: codifies the cross-family rule — with **Mistral** (`vibe -p`, `mistral-medium-3.5`) and **Ornith 35B** (hermes MLX) admitted to verifier duty by verdict parity with the Claude baseline on both canonical verifier tests ([erp#63 evidence](https://gitea.arcodange.lab/arcodange-org/erp/issues/63), blind-judged 16/16 unanimous), the independent verifier SHOULD be a different model family than the builder. - **STATUS truth-pass**: #63 ✅ 2026-07-18 (PR erp#69, harness home `erp:fleet/harness/`), #56 ✅ (PR erp#68 — authored end-to-end by the Mistral builder bench, 0 code corrections), #39 golden set built on local branch `arcodange/golden-set` (push+PR = operator step — blocked by the session permission gate on real financial documents), PR-log rows, P3 → 🟡 in progress. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01VRShc4QhLLU73FLHx9vskh
arcodange added 1 commit 2026-07-18 20:22:33 +02:00
- qa-strategy › Independent verification: with Mistral (vibe -p,
  mistral-medium-3.5) and Ornith 35B admitted to verifier duty by verdict
  parity (erp#63 evidence, blind-judged), the independent verifier SHOULD be
  a different model family than the builder; journal records which family
  verified what.
- STATUS: #63  (PR erp#69, harness home erp:fleet/harness/), #56 
  (PR erp#68, authored by the Mistral builder bench), #39 built on local
  branch (push+PR = operator step), PR-log rows, P3 flipped to in-progress.

Paired with erp#69 (Closes erp#63).

Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01VRShc4QhLLU73FLHx9vskh
arcodange merged commit c8ea130ff6 into main 2026-07-18 20:23:14 +02:00
arcodange deleted branch arcodange/harness-portability-qa 2026-07-18 20:23:14 +02:00
Sign in to join this conversation.
No Reviewers
No labels
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: arcodange-org/factory#31