[Cross-mode] Build quality probe matrix for Ralph and Autoresearch readiness

[devkade]

Objective

Create a repeatable quality surface that reports whether Ralph and Autoresearch are ready across artifacts, workflow, state, runtime, and evidence.

Scope

• CLI/status probe output
• Quality script metrics
• Live QA checklist integration
• Ralph + Autoresearch readiness criteria

Acceptance criteria

• A probe/checklist reports per-mode readiness across artifact completeness, state consistency, event/snapshot parseability, runtime/worktree health, approval state, evidence ledger quality, and terminal closeout conditions.
• Probe output is machine-parseable enough for tests and human-readable enough for operator review.
• Ralph and Autoresearch have separate blocker rollups plus shared cross-mode handoff blockers.
• Probe refuses to hide runtime or semantic blockers behind a high aggregate score.
• Tests cover green, missing artifact, stale state, malformed ledger, missing runtime, missing review evidence, and cross-mode handoff blocker cases.

Quality dimensions

Status visibility, blocker management, operator workflow, regression prevention.

---

Created from Kapi Ralph/Autoresearch quality decomposition requested by Kade.

[devkade]

Assigned to Ragna/coding lane #coding / 7.

• Execution mode: kapi-worker
• Runtime/branch: coding-15 / feat/coding-15
• Worktree/path: /Users/devkade/.kapi/worktrees/coding-15
• Base: origin/dev at 83be567e85a5af82a7ff25a604c0f9ba89fb36e5
• PR if already present: none

Supervisor note: one issue per lane; no destructive cleanup or unrelated refactor. Ragna will verify PR/test/review evidence before claiming completion.