harnessed 3.4.2 → 3.4.4

package/workflows/task/test/SKILL.md

@@ -63,32 +63,19 @@ Phase 01-test 条件性 fire `diagnose` (capabilities.yaml L55-64 mattpocock-ski
 test fail 时进入 diagnose loop (reproduce → minimise → hypothesise → instrument →
 fix → regression-test), 测试通过则 skip diagnose entirely。
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed task-test --task "<text>" --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn + 1-phase (TDD gate evaluation conditional).
-harnessed task-test --task "<text>" --apply
+```bash
+echo "$ARGUMENTS" | harnessed run task-test --task-stdin
 ```
 
-## Forward-looking note
-
-The `trigger_phrases:` frontmatter is active after `harnessed setup` copies this
-SKILL.md to `~/.claude/skills/task-test/` — Claude Code then loads the slash
-command `/task-test` automatically (Gap B fix — sister v1.0.2 mechanism).
-
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run task-test` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.tdd.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/auto/SKILL.md

@@ -62,9 +62,22 @@ Sister `workflows/capabilities.yaml`:
 
 ## Invocation
 
-- CLI: `harnessed verify --phase <num>`
 - Slash command: `/verify` (bare per ADR 0030 namespace policy D-02 LOCK after `harnessed setup`)
 
+## How to invoke
+
+Use the Bash tool to run:
+
+```bash
+echo "$ARGUMENTS" | harnessed run verify --task-stdin
+```
+
+If `$ARGUMENTS` is empty, run `harnessed run verify` (no stdin pipe).
+
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
+
+<!-- harnessed-generated:v3.4.4 -->
+
 ## References
 
 - D-01 master orchestrator delegation pattern

package/workflows/verify/code-review/SKILL.md

@@ -6,7 +6,7 @@ description: |
   多 subagent fan-out, context 隔离, token 敏感)。
   schema_version: harnessed.workflow.v3 with disciplines_applied (6 default) + tools_available
   (code-review) + 1 phase (parallelism ref judgments.parallelism-gate.subagent-default.fires)。
-  Triggered by harnessed CLI `harnessed verify-code-review --phase <num>` or slash command
+  Triggered by slash command
   `/verify-code-review` after `harnessed setup`.
 trigger_phrases:
   - "verify code review"
@@ -48,26 +48,19 @@ Sister `workflows/judgments/parallelism-gate.yaml`:
 总 fire 当 `phase.stage == 'verify'` 后必跑串行 (verify-progress) 之后并行 fan-out。无 skip
 条件 — code-review 多 agent 是 verify-work 第 3 phase 默认 fan-out (sister CLAUDE.md verbatim)。
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-code-review --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn + parallel fan-out subagent.
-harnessed verify-code-review --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-code-review --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-code-review` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.code-review.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/design/SKILL.md

@@ -51,26 +51,19 @@ Sister `workflows/judgments/stage-routing.yaml`:
 - 创意补充 / 不要 AI 味 → `frontend-design`
 - 用户明示「独特 / 不要 AI 感」→ frontend-design 主导, 否则 ui-ux-pro-max 优先
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-design --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn (gate eval true 时).
-harnessed verify-design --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-design --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-design` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.gstack-design-review.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/multispec/SKILL.md

@@ -8,7 +8,7 @@ description: |
   schema_version: harnessed.workflow.v3 with disciplines_applied (6 default) + tools_available
   (agent-teams 3 + 4 specialist capability) + 2 phase (01-team-create on critical-release
   invoke / 02-team-cleanup mandatory shutdown)。
-  Triggered by harnessed CLI `harnessed verify-multispec --phase <num>` or slash command
+  Triggered by slash command
   `/verify-multispec` after `harnessed setup`.
 trigger_phrases:
   - "verify multispec"
@@ -64,26 +64,19 @@ Phase-level `on` clause (critical-release 升级触发):
 - **Token 估算 prereq**: `team_cost < 2 × subagent_cost` (engine-level check per agent-teams.md L34)
 - **Cleanup mandatory**: phase 02-team-cleanup `agent-teams-shutdown` 必跑 (防呆清单)
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-multispec --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn + TeamCreate 4 specialist + 互相 SendMessage 质询 + 末尾 TeamDelete。
-harnessed verify-multispec --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-multispec --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-multispec` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.agent-teams-create.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/paranoid/SKILL.md

@@ -7,7 +7,7 @@ description: |
   默认 critical fire only; 非关键模块 skip (sister CLAUDE.md "关键模块" 限定语)。
   schema_version: harnessed.workflow.v3 with disciplines_applied (6 default) + tools_available
   (gstack-review) + 1 phase (gate ref is_critical_module conditional)。
-  Triggered by harnessed CLI `harnessed verify-paranoid --phase <num>` or slash command
+  Triggered by slash command
   `/verify-paranoid` after `harnessed setup`.
 trigger_phrases:
   - "verify paranoid"
@@ -50,26 +50,19 @@ Sister `workflows/judgments/stage-routing.yaml`:
 - ✅ **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
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-paranoid --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn (gate eval true 时).
-harnessed verify-paranoid --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-paranoid --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-paranoid` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.gstack-review.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/progress/SKILL.md

@@ -46,26 +46,19 @@ Sister `workflows/capabilities.yaml` entries:
 总 fire 当 `phase.stage == 'verify'` (sister `workflows/judgments/stage-routing.yaml`
 verify-progress-always trigger)。无 skip 条件 — verify-work 起点必跑。
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-progress --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn + 3-phase serial chain.
-harnessed verify-progress --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-progress --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-progress` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.gsd-verify-work.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/qa/SKILL.md

@@ -6,7 +6,7 @@ description: |
   schema_version: harnessed.workflow.v3 with disciplines_applied (6 default) + tools_available
   (gstack-qa + playwright-cli + playwright-test + webapp-testing) + 1 phase (gate ref
   has_ui_changes conditional)。
-  Triggered by harnessed CLI `harnessed verify-qa --phase <num>` or slash command
+  Triggered by slash command
   `/verify-qa` after `harnessed setup`.
 trigger_phrases:
   - "verify qa"
@@ -52,26 +52,19 @@ Sister `workflows/judgments/stage-routing.yaml`:
 - setup 需 Python 后端 (Tortoise ORM / pandas) → `webapp-testing` skill
 - 性能 / a11y / 内存诊断 → 不在此 sub-workflow,用 `chrome-devtools-mcp`
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-qa --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn (gate eval true 时).
-harnessed verify-qa --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-qa --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-qa` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.gstack-qa.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/security/SKILL.md

@@ -5,7 +5,7 @@ description: |
   触发, 可选 conditional, sister ~/.claude/CLAUDE.md "Verify 阶段 — 可选 /cso" verbatim)。
   schema_version: harnessed.workflow.v3 with disciplines_applied (6 default) + tools_available
   (gstack-cso) + 1 phase (gate ref has_auth_or_secrets conditional)。
-  Triggered by harnessed CLI `harnessed verify-security --phase <num>` or slash command
+  Triggered by slash command
   `/verify-security` after `harnessed setup`.
 trigger_phrases:
   - "verify security"
@@ -47,26 +47,19 @@ Sister `workflows/judgments/stage-routing.yaml`:
 - ✅ **触发**: auth flow / session / credentials / API keys / SQL injection 路径 / OWASP top 10 area
 - ❌ **跳过**: docs / 纯 UI styling / 内部 refactor / non-security PR
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-security --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn (gate eval true 时).
-harnessed verify-security --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-security --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-security` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.gstack-cso.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/verify/simplify/SKILL.md

@@ -5,7 +5,7 @@ description: |
   sister ~/.claude/CLAUDE.md "Verify 阶段 — 末尾 code-simplifier" verbatim)。
   schema_version: harnessed.workflow.v3 with disciplines_applied (6 default) + tools_available
   (code-simplifier) + 1 phase (gate ref is_final_step 末尾串行)。
-  Triggered by harnessed CLI `harnessed verify-simplify --phase <num>` or slash command
+  Triggered by slash command
   `/verify-simplify` after `harnessed setup`.
 trigger_phrases:
   - "verify simplify"
@@ -47,26 +47,19 @@ Sister `workflows/judgments/stage-routing.yaml`:
 - ✅ **触发**: verify chain 末尾步骤 (所有其他 verify sub 已 ship, 准备 code 简化收尾)
 - ❌ **跳过**: verify chain 中间步骤 (避免过早简化干扰后续 review)
 
-## CLI invocation
+## How to invoke
 
-```bash
-# Dry-run preview — arbitrate-only, never spawns SDK.
-harnessed verify-simplify --phase <num> --dry-run --non-interactive
+Use the Bash tool to run:
 
-# Apply path — real SDK spawn (gate eval true 时).
-harnessed verify-simplify --phase <num> --apply
+```bash
+echo "$ARGUMENTS" | harnessed run verify-simplify --task-stdin
 ```
 
-## How to invoke
+If `$ARGUMENTS` is empty, run `harnessed run verify-simplify` (no stdin pipe).
 
-Use the SlashCommand tool to run: `{{ capabilities.code-simplifier.cmd }}`
+After completion, the Bash output prints a `Next:` hint on stderr suggesting the next stage. Decide whether to invoke based on conversation context — the hint is informational, not prescriptive.
 
-(If a `⚠️ ... not installed` warning was printed by `harnessed setup`, the backing
-capability is missing on disk. Install it (`claude plugin install <name>` for
-plugins, or follow the official install instructions for user-skills — e.g. for
-gstack: `git clone https://github.com/garrytan/gstack.git ~/.claude/skills/gstack` then
-`cd ~/.claude/skills/gstack && ./setup`), then re-run `harnessed setup` to re-render
-this SKILL.md and clear the warning.)
+<!-- harnessed-generated:v3.4.4 -->
 
 ## References
 

package/workflows/execute-task/phases.yaml

@@ -1,73 +0,0 @@
-# workflows/execute-task/phases.yaml — Phase v2.0-2.4 W1.1 v2 schema upgrade
-# Sister v1 (28L, ADR 0011 errata Phase 2.2 W3 T3.3 ship) → v2 (~60L, T2.4.W1.1).
-#
-# v2 delta (per PLAN.md L327-330 verbatim + RESEARCH § 3.3 + D-09 + D-10 + D-13):
-#   - schema_version: harnessed.workflow.v2 (sister Phase 2.2 schemaVersion 7 surface)
-#   - 02-code:    on[] mattpocock route by condition (tdd-gate / grill-with-docs / zoom-out)
-#   - 03-test:    on[] diagnose if test_fail
-#   - 04-deliver: capability + args + gate + parallelism + on[] + fallback (R20.10 explicit halt)
-#
-# Per-phase model tier (ADR 0011 errata / intel CD-2 § 第 4 条) — unchanged from v1.
-# Default table sourced from `.planning/intel/omc-comparison.md` § CD-2 (phase 2.2 W3 T3.3).
-# Override via CLI: `--model-tier inherit` (B-10 escape hatch).
-#
-# capability template — '{{ capabilities.<name>.cmd }}' resolves at workflow runtime
-# (sister D-02 gstack_prefix interpolation pattern from workflows/plan-feature/workflow.yaml).
-# gate / parallelism — 4-level ref 'judgments.<file>.<gate>.fires' pre-resolves via
-# T2.3.W0.4 judgmentResolver BEFORE expr-eval (per src/workflow/schema/workflow.ts L18-19).
-
-schema_version: harnessed.workflow.v2
-workflow: execute-task
-description: 子任务执行 workflow — 4-phase 真接 ralph-loop SDK + tdd-gate conditional + mattpocock route by condition
-phases:
-  - id: 01-clarify
-    name: brainstorming
-    upstream: superpowers brainstorming
-    capability: '{{ capabilities.superpowers-brainstorming.cmd }}'
-    model: opus
-    max_iterations: '{{ defaults.ralph_max_iterations.execute-task.01-clarify }}'
-    gate: judgments.subtask-gate.brainstorming.fires
-
-  - id: 02-code
-    name: code (karpathy 心法 always-on + mattpocock conditional route)
-    upstream: karpathy
-    model: sonnet
-    max_iterations: '{{ defaults.ralph_max_iterations.execute-task.02-code }}'
-    on:
-      - if: 'judgments.tdd-gate.tdd-strongly-suggested.fires'
-        invoke: '{{ capabilities.tdd.cmd }}'
-      - if: 'phase.spec_ambiguous == true'
-        invoke: '{{ capabilities.grill-with-docs.cmd }}'
-      - if: 'phase.unfamiliar_module == true'
-        invoke: '{{ capabilities.zoom-out.cmd }}'
-
-  - id: 03-test
-    name: test (conditional TDD + diagnose on fail)
-    upstream: superpowers TDD
-    capability: '{{ capabilities.tdd.cmd }}'
-    model: sonnet
-    max_iterations: '{{ defaults.ralph_max_iterations.execute-task.03-test }}'
-    on:
-      - if: 'test_fail == true'
-        invoke: '{{ capabilities.diagnose.cmd }}'
-
-  - id: 04-deliver
-    name: deliver (ralph-loop COMPLETE gate + max-iter fallback)
-    upstream: ralph-loop
-    capability: '{{ capabilities.ralph-loop.cmd }}'
-    model: haiku
-    args:
-      completion_promise: COMPLETE
-      max_iterations: '{{ defaults.ralph_max_iterations.execute-task.04-deliver }}'
-    gate: judgments.parallelism-gate.fires
-    parallelism: judgments.parallelism-gate.ralph-loop-wrapper.fires
-    on:
-      - if: 'subtask.lines >= 20 and subtask.type != "single_command_query"'
-        invoke: '{{ capabilities.ralph-loop.cmd }}'
-      - if: 'subtask.lines < 20 or subtask.type == "single_command_query"'
-        action: skip
-    fallback:
-      max_iterations_exceeded:
-        action: emit_warning_and_halt
-        message: '⚠️ ralph-loop max-iterations ({{ args.max_iterations }}) exceeded for execute-task 04-deliver. Sub-task likely incomplete — see workflow engine catch handler for manual options.'
-        exit_code: 1