principles-disciple 1.73.0 → 1.75.0

package/templates/langs/en/principles/THINKING_OS.md

@@ -19,7 +19,7 @@ LLMs are highly sensitive to XML tags; this structure is designed to boost instr
 
   <directive id="T-02" name="PHYSICAL_MEMORY_PERSISTENCE">
     <trigger>When reasoning across multiple files, facing complex debugging, or when the conversation context grows long (>5 turns).</trigger>
-    <must>TRUST FILES, NOT YOUR CONTEXT WINDOW. You MUST actively write your intermediate conclusions, breakpoints, and next steps to `memory/.scratchpad.md` or `PLAN.md`.</must>
+    <must>TRUST FILES, NOT YOUR CONTEXT WINDOW. You MUST actively write your intermediate conclusions, breakpoints, and next steps to `memory/.scratchpad.md`.</must>
     <forbidden>Relying on your internal "brain memory" to hold complex state, which will inevitably be wiped by context compression.</forbidden>
   </directive>
 
@@ -39,8 +39,9 @@ LLMs are highly sensitive to XML tags; this structure is designed to boost instr
   <!-- 执行与物理限制 (Execution & Physical Constraints) -->
   <directive id="T-05" name="PHYSICAL_DEFENSE_AND_ORCHESTRATION">
     <trigger>When asked to perform a major refactoring, multi-file change (>2 files), or an architectural shift.</trigger>
-    <must>Limit your blast radius. You MUST draft a `PLAN.md` (status: READY). After any code change, you MUST run canary tests (e.g., `npm test`, linters) to verify integrity.</must>
-    <forbidden>Executing large-scale unstructured changes directly without a plan, or skipping post-modification validation.</forbidden>
+    <should>For complex changes, describe your plan and get owner confirmation before executing.</should>
+    <must>Limit your blast radius. After any code change, you MUST run canary tests (e.g., `npm test`, linters) to verify integrity.</must>
+    <forbidden>Executing large-scale unstructured changes directly, or skipping post-modification validation.</forbidden>
   </directive>
 
   <directive id="T-06" name="OCCAMS_RAZOR_MVC">

package/templates/langs/en/skills/admin/SKILL.md

@@ -15,7 +15,7 @@ You are now the "Evolutionary System Administrator". Your responsibility is to m
 ### 1. `diagnose` (System Diagnosis)
 **Action**: Check the integrity of the "bare-bones" architecture.
 - **Core Components**: Check if `.claude/hooks/hook_runner.py` exists and is executable.
-- **Documentation Integrity**: Check if `.principles/PROFILE.json`, `PLAN.md` etc. exist.
+- **Documentation Integrity**: Check if `.principles/PROFILE.json` etc. exist.
 - **Tool Awareness**: Check `.state/SYSTEM_CAPABILITIES.json`. If missing, prompt user: "⚠️ Toolchain upgrade not performed. Recommend running `/bootstrap-tools` to significantly enhance system capabilities."
 - **Memory Mount**: Check if `CLAUDE.md` contains `System Integration` section.
 - **Output**: Generate a health report listing missing or abnormal items.
@@ -23,7 +23,7 @@ You are now the "Evolutionary System Administrator". Your responsibility is to m
 ### 2. `repair` (System Repair)
 **Action**: 
 - **Config Recovery**: If `PROFILE.json` is missing or corrupted, attempt recovery from `.claude/templates/PROFILE.json`.
-- **Structure Completion**: Ensure `PLAN.md` contains `## Target Files` heading.
+- **Structure Completion**: Ensure workspace structure is complete.
 - **Forced Cleanup**: Delete `.pain_flag`, `.verdict.json`, `.user_verdict.json`, `.pending_reflection` and other temporary markers.
 
 ### 3. `reset` (Force Reset)

package/templates/langs/en/skills/ai-sprint-orchestration/runtime/.gitignore

@@ -1,2 +1,2 @@
-*
-!.gitignore
+*
+!.gitignore

package/templates/langs/en/skills/evolve-task/SKILL.md

@@ -51,7 +51,7 @@ Output:
 
 ## Step 6: Delegate Planner (Movie Script Plan)
 - Planner outputs Plan (steps/commands/metrics/rollback).
-- Write plan to PLAN.md (STATUS line must exist).
+- Write plan to a planning document for owner review.
 - **Task Sync**: 
   - If `CLAUDE_CODE_TASK_LIST_ID` is set, you must convert the Plan's core steps to Native Tasks (via natural language command "Add task..." or related tools).
   - If not set and in interactive mode, prompt user: "Recommend running `export CLAUDE_CODE_TASK_LIST_ID=task-$(date +%s)` to enable persistent task tracking."
@@ -59,7 +59,7 @@ Output:
 - **Performance Evaluation**: After task completion, write to `.state/.verdict.json`. Format follows `@.principles/schemas/agent_verdict_schema.json`.
 
 ## Step 7: Delegate Implementer (Execution)
-- Implementer can only execute according to PLAN. Any deviation must first update PLAN.
+- Implementer executes according to the plan from Step 6. Any deviation must first be approved by updating the plan.
 - **Performance Evaluation**: After task completion, write to `.state/.verdict.json` based on verification results. Format follows `@.principles/schemas/agent_verdict_schema.json`.
 
 ## Step 8: Delegate Reviewer (Review)

package/templates/langs/en/skills/pd-grooming/SKILL.md

@@ -24,7 +24,7 @@ When performing cleanup operations, you MUST strictly adhere to the following wh
 ### 🌟 Core Assets
 **These files must remain in the root directory. Do not touch:**
 - `AGENTS.md`, `SOUL.md`, `HEARTBEAT.md`, `TOOLS.md`, `IDENTITY.md`, `USER.md`, `MEMORY.md`
-- `README.md`, `PLAN.md`
+- `README.md`
 - `.principles/`, `.state/`
 
 ### 🎯 Targets for Grooming

package/templates/langs/en/skills/pd-mentor/SKILL.md

@@ -185,7 +185,7 @@ For complex scenarios, combine multiple skills:
 
 | Scenario | Combined Flow |
 |----------|---------------|
-| Major refactor | `/pd-evolve` → `plan-script` → `deductive-audit` → execute |
+| Major refactor | `/pd-evolve` → `deductive-audit` → execute |
 | System optimization | `/pd-status` → `evolve-system` → `root-cause` |
 | Project review | `/pd-daily` → `/pd-okr` → `reflection-log` |
 
@@ -196,7 +196,6 @@ These skills are usually called automatically by the system, but advanced users
 - `triage` - Issue triage
 - `root-cause` - Root cause analysis
 - `deductive-audit` - Deductive audit
-- `plan-script` - Plan orchestration
 - `reflection` - Metacognitive reflection
 - `reflection-log` - Reflection logging
 

package/templates/langs/en/skills/reflection/SKILL.md

@@ -12,7 +12,7 @@ disable-model-invocation: false
 Please execute the following reflection steps:
 
 ## 1. Status Scan
-- **Goal**: What was our original objective? (Check `PLAN.md` or early conversation)
+- **Goal**: What was our original objective? (Check early conversation context)
 - **Status**: How much is completed now? Where are we stuck?
 - **Cost**: We've consumed significant tokens. Is the output matching the cost?
 
@@ -37,4 +37,4 @@ If pain detected, must execute:
 
 ## 5. Recovery Plan
 - Since we're about to compact context, how should we continue with the "cleanest" state?
-- Update `PLAN.md`, mark current progress, ensure seamless continuation after compaction.
+- Update `memory/.scratchpad.md`, mark current progress, ensure seamless continuation after compaction.

package/templates/langs/en/skills/report/SKILL.md

@@ -10,4 +10,4 @@ User (the boss) requests an immediate work report.
 
 ## Execution Action
 1. Immediately delegate ``pd_spawn_agent(reporter)``.
-2. Task description: "The boss wants to know the current situation. Please analyze current conversation context, `PLAN.md`, and recent `memory/ISSUE_LOG.md` to write an elegant report for the boss. Remember to check their profile first!"
+2. Task description: "The boss wants to know the current situation. Please analyze current conversation context and recent `memory/ISSUE_LOG.md` to write an elegant report for the boss. Remember to check their profile first!"

package/templates/langs/zh/core/AGENTS.md

@@ -17,7 +17,6 @@
 基于**项目战场**中的相对路径进行决策：
 
 - **项目最高战略**: `./memory/STRATEGY.md`
-- **项目物理计划**: `./PLAN.md`
 - **痛觉反射信号**: Runtime V2 `PainSignalBridge`（手动触发使用 `pd pain record`；`.state/.pain_flag` 仅为 legacy compatibility）
 - **系统能力快照**: `./.state/SYSTEM_CAPABILITIES.json`
 
@@ -158,12 +157,13 @@
 你默认处于架构师模式。
 
 - **L1 (直接执行)**：单文件微调、文档维护 → 直接操作
-- **L2 (委派协议)**：重大变更 → **必须**更新 `./PLAN.md` 并使用 `pd_spawn_agent` 工具委派任务
+- **L2 (委派协议)**：重大变更 → 建议先描述计划并获得 owner 确认后再执行
 
-### 状态机门禁 (State Machine Gating)
+### 计划引导 (Planning Guidance)
 
-- **唯一事实源**：`./PLAN.md`
-- **物理拦截**：插件已激活。若 `PLAN.md` 非 `READY` 且尝试修改风险路径，调用将被阻断
+- 对复杂任务，建议先起草计划文档并获得 owner 批准后再做大幅修改
+- 这是行为建议，不是内置门禁 — PD 默认不强制"先计划后执行"
+- 若 owner 批准的 RuleHost 规则强制了计划行为，该规则会自动生效
 - **防止污染**：禁止将执行层细节写回战略文档
 
 ---

package/templates/langs/zh/core/BOOTSTRAP.md

@@ -183,7 +183,7 @@ openclaw cron add --name "pd-grooming-daily" \
 openclaw cron add --name "health-check" \
   --every 4h \
   --session main \
-  --system-event '环境健康检查：验证核心工具（rg, node, python）是否可用，检查 PLAN.md 状态与实际进度是否一致。'
+  --system-event '环境健康检查：验证核心工具（rg, node, python）是否可用，检查工作区状态与实际进度是否一致。'
 ```
 
 ### 3. 战略对齐检查（每天上午 9 点）

package/templates/langs/zh/principles/THINKING_OS.md

@@ -19,7 +19,7 @@
 
   <directive id="T-02" name="PHYSICAL_MEMORY_PERSISTENCE">
     <trigger>在跨越多个文件进行推理、面临复杂的 Debug、或当对话上下文变得很长（>5 轮）时。</trigger>
-    <must>信任文件，而不是你的上下文窗口。你必须主动将中间结论、断点和后续步骤写入 `memory/.scratchpad.md` 或 `PLAN.md`。</must>
+    <must>信任文件，而不是你的上下文窗口。你必须主动将中间结论、断点和后续步骤写入 `memory/.scratchpad.md`。</must>
     <forbidden>依赖你内部的“大脑记忆”来保持复杂状态，这些状态必然会被上下文压缩机制抹除。</forbidden>
   </directive>
 
@@ -39,8 +39,9 @@
   <!-- 执行与物理限制 (Execution & Physical Constraints) -->
   <directive id="T-05" name="PHYSICAL_DEFENSE_AND_ORCHESTRATION">
     <trigger>当被要求执行大型重构、多文件修改（>2 个文件）或架构变更时。</trigger>
-    <must>限制爆炸半径。你必须起草一个 `PLAN.md`（状态：READY）。在修改任何代码后，必须运行金丝雀测试（例如 `npm test`、linters）以验证完整性。</must>
-    <forbidden>在没有计划的情况下直接执行大规模非结构化变更，或跳过修改后的验证环节。</forbidden>
+    <should>对复杂变更，先描述计划并获得 owner 确认后再执行。</should>
+    <must>限制爆炸半径。在修改任何代码后，必须运行金丝雀测试（例如 `npm test`、linters）以验证完整性。</must>
+    <forbidden>直接执行大规模非结构化变更，或跳过修改后的验证环节。</forbidden>
   </directive>
 
   <directive id="T-06" name="OCCAMS_RAZOR_MVC">

package/templates/langs/zh/skills/admin/SKILL.md

@@ -15,7 +15,7 @@ disable-model-invocation: true
 ### 1. `diagnose` (系统诊断)
 **动作**: 检查“毛坯房”架构的完整性。
 - **核心组件**: 检查 `.claude/hooks/hook_runner.py` 是否存在且可执行。
-- **文档完整性**: 检查 `.principles/PROFILE.json`, `PLAN.md` 等是否存在。
+- **文档完整性**: 检查 `.principles/PROFILE.json` 等是否存在。
 - **工具感知**: 检查 `.state/SYSTEM_CAPABILITIES.json`。若缺失，提示用户："⚠️ 尚未进行工具链升级。建议运行 `/bootstrap-tools` 以大幅提升系统能力。"
 - **记忆挂载**: 检查 `CLAUDE.md` 是否包含 `System Integration` 章节。
 - **输出**: 生成一份健康报告，列出缺失或异常的项目。
@@ -23,7 +23,7 @@ disable-model-invocation: true
 ### 2. `repair` (系统修复)
 **动作**: 
 - **配置恢复**: 如果 `PROFILE.json` 缺失或损坏，尝试从 `.claude/templates/PROFILE.json` 恢复。
-- **结构补全**: 确保 `PLAN.md` 包含 `## Target Files` 标题。
+- **结构补全**: 确保工作区结构完整。
 - **强制清理**: 删除 `.pain_flag`, `.verdict.json`, `.user_verdict.json`, `.pending_reflection` 等临时标记。
 
 ### 3. `reset` (强制重置)

package/templates/langs/zh/skills/ai-sprint-orchestration/runtime/.gitignore

@@ -1,2 +1,2 @@
-*
-!.gitignore
+*
+!.gitignore

package/templates/langs/zh/skills/evolve-task/SKILL.md

@@ -51,7 +51,7 @@ disable-model-invocation: true
 
 ## Step 6: 委派 Planner（电影剧本计划）
 - Planner 输出 Plan（步骤/命令/指标/回滚）。
-- 将计划写入 PLAN.md（STATUS 行必须存在）。
+- 将计划写入计划文档供 owner 审阅。
 - **任务同步 (Task Sync)**: 
   - 如果 `CLAUDE_CODE_TASK_LIST_ID` 已设置，你必须将上述 Plan 的核心步骤直接转化为 Native Tasks（通过自然语言指令"Add task..."或相关工具）。
   - 如果未设置且为交互模式，提示用户："建议运行 `export CLAUDE_CODE_TASK_LIST_ID=task-$(date +%s)` 以启用持久化任务追踪。"
@@ -59,7 +59,7 @@ disable-model-invocation: true
 - **绩效评估**: 任务完成后，写入 `.state/.verdict.json`。格式遵循 `@.principles/schemas/agent_verdict_schema.json`。
 
 ## Step 7: 委派 Implementer（执行）
-- Implementer 只能按 PLAN 执行。任何偏离必须先更新 PLAN。
+- Implementer 按照 Step 6 的计划执行。任何偏离必须先更新计划并获得确认。
 - **绩效评估**: 任务完成后，根据验证结果写入 `.state/.verdict.json`。格式遵循 `@.principles/schemas/agent_verdict_schema.json`。
 
 ## Step 8: 委派 Reviewer（审查）

package/templates/langs/zh/skills/pd-grooming/SKILL.md

@@ -24,7 +24,7 @@ description: 执行工作区"大扫除"，将散落的临时文件归档或清
 ### 🌟 核心资产区 (Core Assets)
 **这些文件必须留在根目录，不可触碰：**
 - `AGENTS.md`, `SOUL.md`, `HEARTBEAT.md`, `TOOLS.md`, `IDENTITY.md`, `USER.md`, `MEMORY.md`
-- `README.md`, `PLAN.md`
+- `README.md`
 - `.principles/`, `.state/`
 
 ### 🎯 可处理区 (Targets for Grooming)

package/templates/langs/zh/skills/pd-mentor/SKILL.md

@@ -185,7 +185,7 @@ disable-model-invocation: true
 
 | 场景 | 组合流程 |
 |------|----------|
-| 大型重构 | `/pd-evolve` → `plan-script` → `deductive-audit` → 执行 |
+| 大型重构 | `/pd-evolve` → `deductive-audit` → 执行 |
 | 系统优化 | `/pd-status` → `evolve-system` → `root-cause` |
 | 项目复盘 | `/pd-daily` → `/pd-okr` → `reflection-log` |
 
@@ -196,7 +196,6 @@ disable-model-invocation: true
 - `triage` - 问题分诊
 - `root-cause` - 根因分析
 - `deductive-audit` - 演绎审计
-- `plan-script` - 计划编排
 - `reflection` - 元认知反思
 - `reflection-log` - 反思落盘
 

package/templates/langs/zh/skills/reflection/SKILL.md

@@ -12,7 +12,7 @@ disable-model-invocation: false
 请执行以下反思步骤：
 
 ## 1. 现状扫描 (Status Scan)
-- **Goal**: 我们最初的目标是什么？(Check `PLAN.md` or early conversation)
+- **Goal**: 我们最初的目标是什么？(检查早期对话上下文)
 - **Status**: 现在完成了多少？卡在哪里？
 - **Cost**: 我们消耗了大量 Token，产出是否匹配？
 
@@ -37,4 +37,4 @@ disable-model-invocation: false
 
 ## 5. 恢复计划 (Recovery)
 - 既然要压缩上下文，我们下一步该如何以“最干净”的状态继续？
-- 更新 `PLAN.md`，标记当前进度，确保压缩后能无缝衔接。
+- 更新 `memory/.scratchpad.md`，标记当前进度，确保压缩后能无缝衔接。

package/templates/langs/zh/skills/report/SKILL.md

@@ -10,4 +10,4 @@ disable-model-invocation: true
 
 ## 执行动作
 1. 立即委派 ``pd_spawn_agent(reporter)``。
-2. 任务描述：“老板想知道当前的情况。请分析当前的对话上下文、`PLAN.md` 和最近的 `memory/ISSUE_LOG.md`，为老板写一份优雅的汇报。请记住先看他的画像！”
+2. 任务描述：“老板想知道当前的情况。请分析当前的对话上下文和最近的 `memory/ISSUE_LOG.md`，为老板写一份优雅的汇报。请记住先看他的画像！”

package/tests/core/migration.test.ts

@@ -18,25 +18,25 @@ describe('Directory Structure Migration', () => {
         vi.clearAllMocks();
     });
 
-    it('should move non-security files from docs/ to new locations', () => {
-        const legacyPlan = path.join(workspaceDir, 'docs', 'PLAN.md');
-        const newPlan = '/mock/workspace/PLAN.md';
+    it('should move THINKING_OS.md from docs/ to .principles/', () => {
+        const legacyThinkingOs = path.join(workspaceDir, 'docs', 'THINKING_OS.md');
+        const newThinkingOs = '/mock/workspace/.principles/THINKING_OS.md';
 
         vi.mocked(fs.existsSync).mockImplementation((p) => {
             const pathStr = p.toString();
             if (pathStr === path.join(workspaceDir, 'docs')) return true;
-            if (pathStr === legacyPlan) return true;
+            if (pathStr === legacyThinkingOs) return true;
             // Destination directories don't exist yet
             if (pathStr === path.join(workspaceDir, '.principles')) return false;
             // Destination files don't exist yet
-            if (pathStr === newPlan) return false;
+            if (pathStr === newThinkingOs) return false;
             return false;
         });
 
         migrateDirectoryStructure(mockApi, workspaceDir);
 
-        // Verify it moved PLAN.md to root
-        expect(fs.renameSync).toHaveBeenCalledWith(legacyPlan, newPlan);
+        // Verify it moved THINKING_OS.md to .principles/
+        expect(fs.renameSync).toHaveBeenCalledWith(legacyThinkingOs, newThinkingOs);
 
         expect(mockLogger.info).toHaveBeenCalledWith(expect.stringContaining('Successfully migrated'));
     });

package/tests/core/path-resolver.test.ts

@@ -20,7 +20,7 @@ describe('PathResolver', () => {
         const { PathResolver } = await import('../../src/core/path-resolver.js');
         const resolver = new PathResolver({ workspaceDir: '/test/workspace' });
         
-        const requiredKeys = ['PROFILE', 'PLAN', 'AGENT_SCORECARD', 'PAIN_FLAG', 'EVOLUTION_QUEUE', 'THINKING_OS', 'THINKING_OS_USAGE', 'THINKING_OS_CANDIDATES'];
+        const requiredKeys = ['PROFILE', 'AGENT_SCORECARD', 'PAIN_FLAG', 'EVOLUTION_QUEUE', 'THINKING_OS', 'THINKING_OS_USAGE', 'THINKING_OS_CANDIDATES'];
         
         for (const key of requiredKeys) {
             expect(() => resolver.resolve(key)).not.toThrow();

package/tests/core/paths-refactor.test.ts

@@ -1,9 +1,5 @@
 import { describe, it, expect, vi, beforeEach } from 'vitest';
-import { planStatus } from '../../src/utils/io.js';
 import { resolvePdPath } from '../../src/core/paths.js';
-import * as fs from 'fs';
-
-vi.mock('fs');
 
 describe('Path Anchoring Integration', () => {
     const workspaceDir = '/mock/workspace';
@@ -17,26 +13,8 @@ describe('Path Anchoring Integration', () => {
         expect(resolvePdPath(workspaceDir, 'PROFILE')).toBe(expected);
     });
 
-    it('should resolve PLAN.md at the project root', () => {
-        const expected = '/mock/workspace/PLAN.md';
-        expect(resolvePdPath(workspaceDir, 'PLAN')).toBe(expected);
-    });
-
     it('should resolve AGENT_SCORECARD.json inside .state/', () => {
         const expected = '/mock/workspace/.state/AGENT_SCORECARD.json';
         expect(resolvePdPath(workspaceDir, 'AGENT_SCORECARD')).toBe(expected);
     });
-
-    it('planStatus should look for PLAN.md in the root', () => {
-        const rootPlanPath = '/mock/workspace/PLAN.md';
-        vi.mocked(fs.existsSync).mockImplementation((p) => p === rootPlanPath);
-        vi.mocked(fs.readFileSync).mockReturnValue('STATUS: READY');
-
-        const status = planStatus(workspaceDir);
-        
-        expect(status).toBe('READY');
-        expect(fs.existsSync).toHaveBeenCalledWith(rootPlanPath);
-        // Verify it does NOT look in docs/
-        expect(fs.existsSync).not.toHaveBeenCalledWith(expect.stringContaining('docs/PLAN.md'));
-    });
 });

package/tests/core/workspace-context.test.ts

@@ -65,8 +65,8 @@ describe('WorkspaceContext', () => {
         
         // PROFILE is at .principles/PROFILE.json
         expect(wctx.resolve('PROFILE')).toBe(path.join(workspaceDir, '.principles', 'PROFILE.json'));
-        // PLAN is at root
-        expect(wctx.resolve('PLAN')).toBe(path.join(workspaceDir, 'PLAN.md'));
+        // THINKING_OS is at .principles/THINKING_OS.md
+        expect(wctx.resolve('THINKING_OS')).toBe(path.join(workspaceDir, '.principles', 'THINKING_OS.md'));
     });
 
     it('should support explicit disposal from cache', () => {

package/tests/core-anti-growth.test.ts

@@ -94,7 +94,6 @@ describe('PRI-212 plugin core anti-growth guard', () => {
     'external-training-contract.ts',
     'merge-gate-audit.ts',
     'shadow-observation-registry.ts',
-    'confirm-first-gate.ts',
     'control-ui-db.ts',
     'thinking-models.ts',
     'pd-task-reconciler.ts',
@@ -123,6 +122,7 @@ describe('PRI-212 plugin core anti-growth guard', () => {
     'evolution-logger.ts',
     'evolution-engine.ts',
     'runtime-v2-prompt-activation-reader.ts',
+    'workspace-guidance-migrator.ts',
   ] as const;
 
   // Category 6: Test files

package/tests/hooks/confirm-first-removal.test.ts

@@ -0,0 +1,188 @@
+/**
+ * PRI-286: Verify confirm-first gate has been fully removed from live paths.
+ *
+ * These tests prove that:
+ * 1. The confirm-first-gate module no longer exists as an importable live module
+ * 2. gate.ts does not call any confirm-first function
+ * 3. prompt.ts does not import any confirm-first function
+ * 4. gate-block-helper does not output confirm-first specific block messages
+ * 5. confirm_first_gate does not appear in DEFAULT_FEATURE_FLAGS
+ * 6. Default PD installation does not block mutating tools due to PLAN.md absence
+ * 7. PLAN.md is not a canonical PD path (paths.ts, path-resolver.ts, env.ts, migration.ts)
+ * 8. confirm-first event types and state store are fully deleted
+ */
+
+import { describe, it, expect } from 'vitest';
+import fs from 'fs';
+import path from 'path';
+
+// __dirname = packages/openclaw-plugin/tests/hooks → go up 4 to monorepo root
+// (hooks → tests → openclaw-plugin → packages → monorepo-root)
+const ROOT = path.resolve(__dirname, '..', '..', '..', '..');
+
+describe('PRI-286: Confirm-first gate removal verification', () => {
+  it('confirm-first-gate.ts source file has been deleted', () => {
+    const gatePath = path.join(ROOT, 'packages/openclaw-plugin/src/core/confirm-first-gate.ts');
+    expect(fs.existsSync(gatePath)).toBe(false);
+  });
+
+  it('gate.ts does not import from confirm-first-gate', async () => {
+    const gateSource = fs.readFileSync(
+      path.join(ROOT, 'packages/openclaw-plugin/src/hooks/gate.ts'),
+      'utf8',
+    );
+    expect(gateSource).not.toContain('confirm-first-gate');
+    expect(gateSource).not.toContain('evaluateConfirmFirstGateSync');
+    expect(gateSource).not.toContain('confirm-first');
+  });
+
+  it('prompt.ts does not import from confirm-first-gate', async () => {
+    const promptSource = fs.readFileSync(
+      path.join(ROOT, 'packages/openclaw-plugin/src/hooks/prompt.ts'),
+      'utf8',
+    );
+    expect(promptSource).not.toContain('confirm-first-gate');
+    expect(promptSource).not.toContain('detectApprovalMarker');
+    expect(promptSource).not.toContain('setConfirmFirstDirective');
+    expect(promptSource).not.toContain('setConfirmFirstApproval');
+    expect(promptSource).not.toContain('hydrateFromStore');
+    expect(promptSource).not.toContain('pruneStoreStaleRows');
+    expect(promptSource).not.toContain('setConfirmFirstStore');
+    expect(promptSource).not.toContain('resetConfirmFirst');
+    expect(promptSource).not.toContain('setConfirmFirstGateEnabled');
+    expect(promptSource).not.toContain('SqliteConfirmFirstStateStore');
+    expect(promptSource).not.toContain('confirm_first_gate');
+  });
+
+  it('gate-block-helper does not have confirm-first specific branch', () => {
+    const helperSource = fs.readFileSync(
+      path.join(ROOT, 'packages/openclaw-plugin/src/hooks/gate-block-helper.ts'),
+      'utf8',
+    );
+    expect(helperSource).not.toContain('confirm-first-gate');
+    expect(helperSource).not.toContain('Confirm-First Gate Blocked');
+    expect(helperSource).not.toContain('confirm-first behavioral directive');
+  });
+
+  it('confirm_first_gate is not in DEFAULT_FEATURE_FLAGS', async () => {
+    const { DEFAULT_FEATURE_FLAGS } = await import('@principles/core/runtime-v2');
+    const ids = DEFAULT_FEATURE_FLAGS.map((f: { id: string }) => f.id);
+    expect(ids).not.toContain('confirm_first_gate');
+  });
+
+  it('no PLAN.md physical interception language in AGENTS.md templates', () => {
+    const templateDirs = [
+      path.join(ROOT, 'packages/openclaw-plugin/templates'),
+      path.join(ROOT, 'packages/create-principles-disciple/templates'),
+      path.join(ROOT, 'packages/create-principles-disciple/plugin/templates'),
+    ];
+
+    for (const dir of templateDirs) {
+      if (!fs.existsSync(dir)) continue;
+      const agentsFiles = findFiles(dir, 'AGENTS.md');
+      for (const file of agentsFiles) {
+        const content = fs.readFileSync(file, 'utf8');
+        // Must NOT contain physical interception language
+        expect(content, `${file} should not contain physical interception`).not.toContain('Physical interception');
+        expect(content, `${file} should not contain 物理拦截`).not.toContain('物理拦截');
+        expect(content, `${file} should not contain Single source of truth.*PLAN`).not.toMatch(/Single source of truth.*PLAN/i);
+      }
+    }
+  });
+
+  it('no mandatory PLAN.md STATUS:READY in THINKING_OS templates', () => {
+    const templateDirs = [
+      path.join(ROOT, 'packages/openclaw-plugin/templates'),
+      path.join(ROOT, 'packages/create-principles-disciple/templates'),
+      path.join(ROOT, 'packages/create-principles-disciple/plugin/templates'),
+    ];
+
+    for (const dir of templateDirs) {
+      if (!fs.existsSync(dir)) continue;
+      const thinkingFiles = findFiles(dir, 'THINKING_OS.md');
+      for (const file of thinkingFiles) {
+        const content = fs.readFileSync(file, 'utf8');
+        expect(content, `${file} should not require PLAN.md status: READY`).not.toContain('PLAN.md` (status: READY)');
+        expect(content, `${file} should not require PLAN.md（状态：READY）`).not.toContain('PLAN.md`（状态：READY）');
+      }
+    }
+  });
+
+  // ── Round 2: Canonical PLAN.md path removal ──
+
+  it('paths.ts does not contain PLAN: entry', () => {
+    const source = fs.readFileSync(
+      path.join(ROOT, 'packages/openclaw-plugin/src/core/paths.ts'),
+      'utf8',
+    );
+    expect(source).not.toMatch(/PLAN:\s*'PLAN\.md'/);
+  });
+
+  it('path-resolver.ts does not contain PLAN key', () => {
+    const source = fs.readFileSync(
+      path.join(ROOT, 'packages/openclaw-plugin/src/core/path-resolver.ts'),
+      'utf8',
+    );
+    expect(source).not.toContain("'PLAN':");
+    expect(source).not.toContain('"PLAN":');
+  });
+
+  it('env.ts CORE_FILES does not contain PLAN.md', () => {
+    const source = fs.readFileSync(
+      path.join(ROOT, 'packages/create-principles-disciple/src/utils/env.ts'),
+      'utf8',
+    );
+    // Match 'PLAN.md' inside the CORE_FILES array — should not exist
+    expect(source).not.toMatch(/CORE_FILES\s*=\s*\[[\s\S]*?'PLAN\.md'/);
+  });
+
+  it('migration.ts does not migrate docs/PLAN.md', () => {
+    const source = fs.readFileSync(
+      path.join(ROOT, 'packages/openclaw-plugin/src/core/migration.ts'),
+      'utf8',
+    );
+    expect(source).not.toContain("'PLAN.md'");
+    expect(source).not.toContain("newKey: 'PLAN'");
+  });
+
+  // ── Round 2: Event type and state store full deletion ──
+
+  it('event-types.ts does not contain runtime_v2_confirm_first_gate', () => {
+    const source = fs.readFileSync(
+      path.join(ROOT, 'packages/principles-core/src/runtime-v2/types/event-types.ts'),
+      'utf8',
+    );
+    expect(source).not.toContain('runtime_v2_confirm_first_gate');
+    expect(source).not.toContain('RuntimeV2ConfirmFirstGate');
+  });
+
+  it('confirm-first state store source has been deleted', () => {
+    const storePath = path.join(
+      ROOT, 'packages/principles-core/src/runtime-v2/activation/sqlite-confirm-first-state-store.ts',
+    );
+    expect(fs.existsSync(storePath)).toBe(false);
+  });
+
+  it('confirm-first state store test has been deleted', () => {
+    const testPath = path.join(
+      ROOT, 'packages/principles-core/src/runtime-v2/__tests__/sqlite-confirm-first-state-store.test.ts',
+    );
+    expect(fs.existsSync(testPath)).toBe(false);
+  });
+});
+
+function findFiles(dir: string, filename: string): string[] {
+  const results: string[] = [];
+  function walk(d: string): void {
+    for (const entry of fs.readdirSync(d, { withFileTypes: true })) {
+      const full = path.join(d, entry.name);
+      if (entry.isDirectory()) {
+        walk(full);
+      } else if (entry.name === filename) {
+        results.push(full);
+      }
+    }
+  }
+  walk(dir);
+  return results;
+}