Paranoid

verify-paranoid workflow (v3)

Overview

1-phase sub-workflow mapping CLAUDE.md "gstack 治理关卡 🔒 关键模块 PR 前强制 — /review" onto harnessed runtime (Phase v3.0-3.4 W0.12 — D-04 Stage ④ Verify 7 sub + D-12 gstack 治理关卡 + Pattern A sub-workflow ship)。

phase	id	upstream	model	capability	gate
1	01-review	gstack	opus	{{ capabilities.gstack-review.cmd }}	judgments.stage-routing.verify-paranoid-critical.fires

Per-phase config loads from workflows/verify/paranoid/workflow.yaml; engine 4-level gate resolver evaluates phase.is_critical_module == true via expr-eval — true 则 invoke gstack /review, false 则 skip (chain_isolation 3 铁律 R20.16 sister verify-work v2 phase 04)。

Capability refs

Sister workflows/capabilities.yaml entries:

• gstack-review — Bucket 3 治理关卡 (impl: gstack, cmd: /review, fires_when: phase.is_critical_module == true)

Gate ref

Sister workflows/judgments/stage-routing.yaml:

• verify-paranoid-critical.fires — phase.stage == 'verify' and phase.is_critical_module == true (默认 critical fire only; 普通 PR 应跳过 — gstack-review 是 Paranoid Staff Engineer 重武器)

Routing rules (sister CLAUDE.md "gstack 治理关卡")

• ✅ 触发: 关键模块 PR 前 (auth / payment / data migration / core algorithm 等)
• ❌ 跳过: 常规 PR / docs / config / 非核心 module

How to invoke

!harnessed checkpoint intent verify-paranoid

The banner above (when present) means this invocation is REGISTERED with the engine (an intent marker) — not yet compliant: the steps below (prompt → spawn → checkpoint complete) resolve it, and a per-turn <workflow-intent> reminder persists until they run.

The numbered sequence below is the state machine — execute it with Bash. Do NOT improvise an equivalent flow from the Overview above: freelancing bypasses the engine (no ledger, no evidence guard). harnessed gives you the spawn-ready prompt; YOU spawn the subagent with a CC-native Task / Agent tool (keeps the session responsive + lets clarification round-trips reach the user).

Do NOT pipe to harnessed run verify-paranoid — that is the CI/headless path (in-process SDK spawn that blocks the session inside Claude Code).

1. Bash: harnessed prompt verify-paranoid --task "$ARGUMENTS" --json → parse {prompt, max_iterations, model}.
2. Spawn a CC-native subagent (Task / Agent tool) with that prompt + model, wrapped in the ralph-loop plugin: /ralph-loop "<prompt>" --max-iterations <max_iterations> --completion-promise "COMPLETE". If the plugin is absent, use the native goal gate instead (Claude Code 2.1.139+ / Codex): /goal "this subtask is delivered: the subagent's final output contains verbatim <promise>COMPLETE</promise>; or stop after <max_iterations> turns" then spawn the subagent and let the goal evaluator drive re-spawns until it clears. If /goal is unavailable too, self-loop: spawn → check output for <promise>COMPLETE</promise> → re-spawn with prior output appended (up to max_iterations). Set the goal only at the leaf subtask level — /goal is single-slot per session and a nested goal overwrites the outer one.
3. If the output contains STATUS: NEEDS_CLARIFICATION + a question list: STOP, relay them verbatim via AskUserQuestion, append the answers to the spec, then re-spawn the same sub.
4. On <promise>COMPLETE</promise>: Bash harnessed checkpoint complete verify-paranoid --summary "<one-line>". The evidence guard runs here (fail-CLOSED): if a declared artifacts_expected file is missing it exits non-zero — re-spawn to produce it before treating the sub as done.

<!-- harnessed-generated:v4.9.3 -->

References

• D-04 Stage ④ Verify 7 sub 分解
• D-12 gstack 治理关卡强制
• workflows/capabilities.yaml — gstack-review
• workflows/judgments/stage-routing.yaml — verify-paranoid-critical trigger
• workflows/defaults.yaml — ralph_max_iterations.verify-paranoid.* values (W2.2 backfill)
• workflows/verify-work/workflow.yaml v2 SHIPPED phase 04-gstack-review-conditional sister verbatim