principles-disciple 1.209.4 → 1.209.6
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +73 -0
- package/dist/bundle.js +13 -11
- package/dist/constants/tools.d.ts +1 -1
- package/dist/hooks/prompt-helpers.d.ts +5 -5
- package/dist/openclaw.plugin.json +1 -1
- package/dist/templates/langs/en/core/AGENTS.md +40 -180
- package/dist/templates/langs/en/core/BOOT.md +19 -47
- package/dist/templates/langs/en/core/HEARTBEAT.md +26 -58
- package/dist/templates/langs/en/core/IDENTITY.md +15 -8
- package/dist/templates/langs/en/core/PRINCIPLES.md +6 -7
- package/dist/templates/langs/en/core/SOUL.md +10 -41
- package/dist/templates/langs/en/core/TOOLS.md +18 -10
- package/dist/templates/langs/en/core/USER.md +13 -8
- package/dist/templates/langs/en/skills/pd-auditor/SKILL.md +1 -1
- package/dist/templates/langs/en/skills/pd-explorer/SKILL.md +1 -1
- package/dist/templates/langs/en/skills/pd-implementer/SKILL.md +1 -1
- package/dist/templates/langs/en/skills/pd-mentor/SKILL.md +6 -30
- package/dist/templates/langs/en/skills/pd-planner/SKILL.md +1 -1
- package/dist/templates/langs/zh/core/AGENTS.md +38 -173
- package/dist/templates/langs/zh/core/BOOT.md +14 -46
- package/dist/templates/langs/zh/core/HEARTBEAT.md +20 -82
- package/dist/templates/langs/zh/core/IDENTITY.md +15 -8
- package/dist/templates/langs/zh/core/PRINCIPLES.md +7 -7
- package/dist/templates/langs/zh/core/SOUL.md +9 -41
- package/dist/templates/langs/zh/core/TOOLS.md +14 -10
- package/dist/templates/langs/zh/core/USER.md +12 -8
- package/dist/templates/langs/zh/skills/pd-auditor/SKILL.md +1 -1
- package/dist/templates/langs/zh/skills/pd-explorer/SKILL.md +1 -1
- package/dist/templates/langs/zh/skills/pd-implementer/SKILL.md +1 -1
- package/dist/templates/langs/zh/skills/pd-mentor/SKILL.md +6 -30
- package/dist/templates/langs/zh/skills/pd-planner/SKILL.md +1 -1
- package/dist/templates/pain_settings.json +1 -2
- package/openclaw.plugin.json +1 -1
- package/package.json +1 -2
- package/templates/langs/en/core/AGENTS.md +40 -180
- package/templates/langs/en/core/BOOT.md +19 -47
- package/templates/langs/en/core/HEARTBEAT.md +26 -58
- package/templates/langs/en/core/IDENTITY.md +15 -8
- package/templates/langs/en/core/PRINCIPLES.md +6 -7
- package/templates/langs/en/core/SOUL.md +10 -41
- package/templates/langs/en/core/TOOLS.md +18 -10
- package/templates/langs/en/core/USER.md +13 -8
- package/templates/langs/en/skills/pd-auditor/SKILL.md +1 -1
- package/templates/langs/en/skills/pd-explorer/SKILL.md +1 -1
- package/templates/langs/en/skills/pd-implementer/SKILL.md +1 -1
- package/templates/langs/en/skills/pd-mentor/SKILL.md +6 -30
- package/templates/langs/en/skills/pd-planner/SKILL.md +1 -1
- package/templates/langs/zh/core/AGENTS.md +38 -173
- package/templates/langs/zh/core/BOOT.md +14 -46
- package/templates/langs/zh/core/HEARTBEAT.md +20 -82
- package/templates/langs/zh/core/IDENTITY.md +15 -8
- package/templates/langs/zh/core/PRINCIPLES.md +7 -7
- package/templates/langs/zh/core/SOUL.md +9 -41
- package/templates/langs/zh/core/TOOLS.md +14 -10
- package/templates/langs/zh/core/USER.md +12 -8
- package/templates/langs/zh/skills/pd-auditor/SKILL.md +1 -1
- package/templates/langs/zh/skills/pd-explorer/SKILL.md +1 -1
- package/templates/langs/zh/skills/pd-implementer/SKILL.md +1 -1
- package/templates/langs/zh/skills/pd-mentor/SKILL.md +6 -30
- package/templates/langs/zh/skills/pd-planner/SKILL.md +1 -1
- package/templates/pain_settings.json +1 -2
- package/dist/templates/langs/en/core/BOOTSTRAP.md +0 -259
- package/dist/templates/langs/en/skills/admin/SKILL.md +0 -39
- package/dist/templates/langs/en/skills/bootstrap-tools/SKILL.md +0 -53
- package/dist/templates/langs/en/skills/deductive-audit/SKILL.md +0 -36
- package/dist/templates/langs/en/skills/evolution-framework-update/SKILL.md +0 -31
- package/dist/templates/langs/en/skills/evolve-task/SKILL.md +0 -83
- package/dist/templates/langs/en/skills/feedback/SKILL.md +0 -51
- package/dist/templates/langs/en/skills/init-strategy/SKILL.md +0 -54
- package/dist/templates/langs/en/skills/inject-rule/SKILL.md +0 -19
- package/dist/templates/langs/en/skills/profile/SKILL.md +0 -24
- package/dist/templates/langs/en/skills/reflection/SKILL.md +0 -40
- package/dist/templates/langs/en/skills/reflection-log/SKILL.md +0 -37
- package/dist/templates/langs/en/skills/report/SKILL.md +0 -13
- package/dist/templates/langs/en/skills/root-cause/SKILL.md +0 -33
- package/dist/templates/langs/en/skills/triage/SKILL.md +0 -29
- package/dist/templates/langs/en/skills/watch-evolution/SKILL.md +0 -33
- package/dist/templates/langs/zh/core/BOOTSTRAP.md +0 -260
- package/dist/templates/langs/zh/skills/admin/SKILL.md +0 -41
- package/dist/templates/langs/zh/skills/bootstrap-tools/SKILL.md +0 -52
- package/dist/templates/langs/zh/skills/deductive-audit/SKILL.md +0 -36
- package/dist/templates/langs/zh/skills/evolution-framework-update/SKILL.md +0 -31
- package/dist/templates/langs/zh/skills/evolve-task/SKILL.md +0 -83
- package/dist/templates/langs/zh/skills/feedback/SKILL.md +0 -53
- package/dist/templates/langs/zh/skills/init-strategy/SKILL.md +0 -54
- package/dist/templates/langs/zh/skills/inject-rule/SKILL.md +0 -19
- package/dist/templates/langs/zh/skills/profile/SKILL.md +0 -24
- package/dist/templates/langs/zh/skills/reflection/SKILL.md +0 -40
- package/dist/templates/langs/zh/skills/reflection-log/SKILL.md +0 -37
- package/dist/templates/langs/zh/skills/report/SKILL.md +0 -13
- package/dist/templates/langs/zh/skills/root-cause/SKILL.md +0 -33
- package/dist/templates/langs/zh/skills/triage/SKILL.md +0 -29
- package/dist/templates/langs/zh/skills/watch-evolution/SKILL.md +0 -33
- package/scripts/acceptance-test.mjs +0 -257
- package/scripts/bootstrap-rules.mjs +0 -70
- package/scripts/compile-principles.mjs +0 -104
- package/scripts/install-dependencies.cjs +0 -47
- package/scripts/postinstall.cjs +0 -177
- package/scripts/sync-plugin.mjs +0 -1442
- package/scripts/verify-build.mjs +0 -120
- package/templates/langs/en/core/BOOTSTRAP.md +0 -259
- package/templates/langs/en/skills/admin/SKILL.md +0 -39
- package/templates/langs/en/skills/bootstrap-tools/SKILL.md +0 -53
- package/templates/langs/en/skills/deductive-audit/SKILL.md +0 -36
- package/templates/langs/en/skills/evolution-framework-update/SKILL.md +0 -31
- package/templates/langs/en/skills/evolve-task/SKILL.md +0 -83
- package/templates/langs/en/skills/feedback/SKILL.md +0 -51
- package/templates/langs/en/skills/init-strategy/SKILL.md +0 -54
- package/templates/langs/en/skills/inject-rule/SKILL.md +0 -19
- package/templates/langs/en/skills/profile/SKILL.md +0 -24
- package/templates/langs/en/skills/reflection/SKILL.md +0 -40
- package/templates/langs/en/skills/reflection-log/SKILL.md +0 -37
- package/templates/langs/en/skills/report/SKILL.md +0 -13
- package/templates/langs/en/skills/root-cause/SKILL.md +0 -33
- package/templates/langs/en/skills/triage/SKILL.md +0 -29
- package/templates/langs/en/skills/watch-evolution/SKILL.md +0 -33
- package/templates/langs/zh/core/BOOTSTRAP.md +0 -260
- package/templates/langs/zh/skills/admin/SKILL.md +0 -41
- package/templates/langs/zh/skills/bootstrap-tools/SKILL.md +0 -52
- package/templates/langs/zh/skills/deductive-audit/SKILL.md +0 -36
- package/templates/langs/zh/skills/evolution-framework-update/SKILL.md +0 -31
- package/templates/langs/zh/skills/evolve-task/SKILL.md +0 -83
- package/templates/langs/zh/skills/feedback/SKILL.md +0 -53
- package/templates/langs/zh/skills/init-strategy/SKILL.md +0 -54
- package/templates/langs/zh/skills/inject-rule/SKILL.md +0 -19
- package/templates/langs/zh/skills/profile/SKILL.md +0 -24
- package/templates/langs/zh/skills/reflection/SKILL.md +0 -40
- package/templates/langs/zh/skills/reflection-log/SKILL.md +0 -37
- package/templates/langs/zh/skills/report/SKILL.md +0 -13
- package/templates/langs/zh/skills/root-cause/SKILL.md +0 -33
- package/templates/langs/zh/skills/triage/SKILL.md +0 -29
- package/templates/langs/zh/skills/watch-evolution/SKILL.md +0 -33
|
@@ -1,15 +1,23 @@
|
|
|
1
|
-
#
|
|
1
|
+
# TOOLS.md - Tool Conventions
|
|
2
2
|
|
|
3
|
-
|
|
4
|
-
- **Map First**: Before any file search, **must** first consult architecture diagrams or code maps under `docs/`.
|
|
5
|
-
- **Deterministic Execution**: Before writing code, must achieve 100% context certainty. No guessing-based programming.
|
|
6
|
-
- **Tool Preference**: Prefer `rg` (ripgrep) for high-performance search. Never blindly traverse.
|
|
3
|
+
This file is used by the host/OpenClaw for local tool conventions.
|
|
7
4
|
|
|
5
|
+
Principles Disciple only adds conventions for its own PD-specific commands.
|
|
6
|
+
Everything else (editor habits, search preferences, shell workflows) is up to
|
|
7
|
+
you — PD does not prescribe general coding behavior.
|
|
8
8
|
|
|
9
|
+
---
|
|
9
10
|
|
|
10
|
-
##
|
|
11
|
+
## PD-specific commands
|
|
11
12
|
|
|
12
|
-
-
|
|
13
|
-
-
|
|
14
|
-
-
|
|
15
|
-
|
|
13
|
+
- **Record behavior evidence**: `pd pain record` — manual trigger for
|
|
14
|
+
owner-relevant behavior evidence (not every tool failure).
|
|
15
|
+
- **Review principle proposals**: `pd candidate list` — proposals awaiting
|
|
16
|
+
Owner decision (approve / reject / defer / rollback).
|
|
17
|
+
- **Owner console**: `pd console open` — visual review and rollback controls.
|
|
18
|
+
- **RuleHost / code_tool_hook activation**: managed via `pd` — the hard-gate
|
|
19
|
+
activation channel for approved principles.
|
|
20
|
+
|
|
21
|
+
---
|
|
22
|
+
|
|
23
|
+
_Add your own tool conventions here._
|
|
@@ -1,10 +1,15 @@
|
|
|
1
|
-
#
|
|
1
|
+
# USER.md - User Context Placeholder
|
|
2
2
|
|
|
3
|
-
|
|
4
|
-
- **Preference**: User prefers English communication.
|
|
5
|
-
- **Expertise**: Adjust your guidance intensity dynamically based on `USER_PROFILE.json`.
|
|
6
|
-
- **Vision**: Build a top-tier agent framework capable of self-repair and self-evolution.
|
|
3
|
+
This file is used by the host/OpenClaw to understand who the user is.
|
|
7
4
|
|
|
8
|
-
|
|
9
|
-
|
|
10
|
-
|
|
5
|
+
Principles Disciple does **not** preset any real user preference, expertise,
|
|
6
|
+
or vision, and it does not ship automatic behaviors tied to user profiles.
|
|
7
|
+
Fill this file in with your own context, or leave the placeholder as-is.
|
|
8
|
+
|
|
9
|
+
<!-- Replace with your own context, for example:
|
|
10
|
+
|
|
11
|
+
- Preferred language: <your preference>
|
|
12
|
+
- How you like to collaborate: <notes>
|
|
13
|
+
- Anything else the agent should know about you
|
|
14
|
+
|
|
15
|
+
-->
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: pd-auditor
|
|
3
|
-
description: Deductive audit
|
|
3
|
+
description: Deductive audit of target systems within Principles Disciple governance scope: axiom verification, system audit, and via negativa. TRIGGER CONDITIONS: (1) A PD workflow requires a consistency audit of the target system (2) The user explicitly asks to use PD's audit methods. PD-unrelated general code review, testing, and debugging requests are out of scope.
|
|
4
4
|
disable-model-invocation: true
|
|
5
5
|
---
|
|
6
6
|
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: pd-explorer
|
|
3
|
-
description: Rapid evidence collection
|
|
3
|
+
description: Rapid evidence collection for Principles Disciple pain diagnosis: locate files, logs, and reproduction steps. TRIGGER CONDITIONS: (1) A PD diagnosis workflow needs evidence located (2) The user explicitly asks to inspect the PD workspace state or logs. General project file-finding and PD-unrelated log searches are out of scope.
|
|
4
4
|
disable-model-invocation: true
|
|
5
5
|
---
|
|
6
6
|
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: pd-implementer
|
|
3
|
-
description: Execute code modifications according to
|
|
3
|
+
description: Execute code modifications according to an approved plan inside Principles Disciple workflows. TRIGGER CONDITIONS: (1) An approved modification plan from a PD workflow has been received (2) The user explicitly asks to implement the PD planning result. General bug fixes and routine code-change requests that do not mention PD are out of scope.
|
|
4
4
|
disable-model-invocation: true
|
|
5
5
|
---
|
|
6
6
|
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: pd-mentor
|
|
3
|
-
description: Interactive command guidance and scenario-based recommendations
|
|
3
|
+
description: Interactive command guidance and scenario-based recommendations for using Principles Disciple itself. TRIGGER CONDITIONS: (1) The user explicitly asks about Principles Disciple / PD usage, commands, configuration, or workflows (2) The user explicitly asks to use PD built-in capabilities for a governance goal. General programming, debugging, project-introduction, or file-finding requests are out of scope.
|
|
4
4
|
disable-model-invocation: true
|
|
5
5
|
---
|
|
6
6
|
|
|
@@ -39,7 +39,7 @@ I'm your intelligent mentor, helping you understand and use all features of Prin
|
|
|
39
39
|
|
|
40
40
|
### Scenario 1: New Project Initialization
|
|
41
41
|
|
|
42
|
-
**Trigger**:
|
|
42
|
+
**Trigger**: The user explicitly mentions onboarding a new project with Principles Disciple and asks for PD setup guidance
|
|
43
43
|
|
|
44
44
|
**Recommended Flow**:
|
|
45
45
|
1. `/pd-init` - Establish strategic vision
|
|
@@ -52,7 +52,7 @@ I'm your intelligent mentor, helping you understand and use all features of Prin
|
|
|
52
52
|
|
|
53
53
|
### Scenario 2: Bug Fix Needed
|
|
54
54
|
|
|
55
|
-
**Trigger**:
|
|
55
|
+
**Trigger**: The user explicitly reports a problem with Principles Disciple / PD itself (a PD command failing, runtime status issues, security-stage questions)
|
|
56
56
|
|
|
57
57
|
**Recommended Flow**:
|
|
58
58
|
1. `/pd-status` - Check system status (GFI and Pain Dictionary)
|
|
@@ -64,7 +64,7 @@ I'm your intelligent mentor, helping you understand and use all features of Prin
|
|
|
64
64
|
|
|
65
65
|
### Scenario 3: Daily Maintenance & Review
|
|
66
66
|
|
|
67
|
-
**Trigger**:
|
|
67
|
+
**Trigger**: The user explicitly asks for a Principles Disciple progress or health review (mentioning PD, EP, GFI, or system status)
|
|
68
68
|
|
|
69
69
|
**Recommended Flow**:
|
|
70
70
|
1. `/pd-evolution-status` - View current trust score
|
|
@@ -76,7 +76,7 @@ I'm your intelligent mentor, helping you understand and use all features of Prin
|
|
|
76
76
|
|
|
77
77
|
### Scenario 4: Permission or Security Related
|
|
78
78
|
|
|
79
|
-
**Trigger**:
|
|
79
|
+
**Trigger**: The user asks about Principles Disciple permissions, blocks, or security stages
|
|
80
80
|
|
|
81
81
|
**Recommended Flow**:
|
|
82
82
|
1. `/pd-evolution-status` - View trust score and security stage
|
|
@@ -88,7 +88,7 @@ I'm your intelligent mentor, helping you understand and use all features of Prin
|
|
|
88
88
|
|
|
89
89
|
### Scenario 6: Tool Upgrade Needs
|
|
90
90
|
|
|
91
|
-
**Trigger**:
|
|
91
|
+
**Trigger**: The user explicitly asks to use PD's environment scan or tool-research capabilities
|
|
92
92
|
|
|
93
93
|
**Recommended Flow**:
|
|
94
94
|
1. `/pd-bootstrap` - Scan current environment
|
|
@@ -161,30 +161,6 @@ Ask user if they want to execute a command:
|
|
|
161
161
|
|
|
162
162
|
---
|
|
163
163
|
|
|
164
|
-
## Advanced Usage
|
|
165
|
-
|
|
166
|
-
### Combined Skills
|
|
167
|
-
|
|
168
|
-
For complex scenarios, combine multiple skills:
|
|
169
|
-
|
|
170
|
-
| Scenario | Combined Flow |
|
|
171
|
-
|----------|---------------|
|
|
172
|
-
| Major refactor | `/pd-evolution-status` → `deductive-audit` → execute |
|
|
173
|
-
| System optimization | `/pd-status` → `root-cause` → optimize |
|
|
174
|
-
| Project review | `/pd-evolution-status` → `/pd-status` → `reflection-log` |
|
|
175
|
-
|
|
176
|
-
### Internal Skill Calls
|
|
177
|
-
|
|
178
|
-
These skills are usually called automatically by the system, but advanced users can also trigger them manually:
|
|
179
|
-
|
|
180
|
-
- `triage` - Issue triage
|
|
181
|
-
- `root-cause` - Root cause analysis
|
|
182
|
-
- `deductive-audit` - Deductive audit
|
|
183
|
-
- `reflection` - Metacognitive reflection
|
|
184
|
-
- `reflection-log` - Reflection logging
|
|
185
|
-
|
|
186
|
-
---
|
|
187
|
-
|
|
188
164
|
## FAQ
|
|
189
165
|
|
|
190
166
|
**Q: What is GFI (Friction Index)?**
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: pd-planner
|
|
3
|
-
description: Movie-script-style planning
|
|
3
|
+
description: Movie-script-style planning inside Principles Disciple workflows: break an approved complex task into executable steps. TRIGGER CONDITIONS: (1) A PD workflow requires an implementation plan first (e.g. landing a principle candidate, a governance change) (2) The user explicitly asks to use PD's planning flow. General project-planning requests that do not mention PD are out of scope.
|
|
4
4
|
disable-model-invocation: true
|
|
5
5
|
---
|
|
6
6
|
|
|
@@ -1,197 +1,62 @@
|
|
|
1
|
-
#
|
|
1
|
+
# AGENTS.md - 工作空间指南
|
|
2
2
|
|
|
3
|
-
|
|
3
|
+
Principles Disciple(PD)是 **Owner 治理的行为内化层**:把重复出现的、与 Owner 相关的
|
|
4
|
+
行为证据,转化为经过审阅、可回滚的原则,从而改变 Agent 未来的行为。
|
|
4
5
|
|
|
5
|
-
|
|
6
|
-
|
|
7
|
-
1. **中枢神经 (Agent Workspace)**:
|
|
8
|
-
- 存放核心 DNA (`SOUL.md`, `AGENTS.md`) 的目录
|
|
9
|
-
- 属于"意识空间",严禁将项目业务逻辑写入此处
|
|
10
|
-
|
|
11
|
-
2. **项目战场 (Project Root)**:
|
|
12
|
-
- 当前执行命令的工作目录 (`$CWD`)
|
|
13
|
-
- 存放业务代码 (`src/`)、项目文档 (`docs/`) 和战略资产
|
|
14
|
-
|
|
15
|
-
## 🎯 核心事实源 (Truth Anchors)
|
|
16
|
-
|
|
17
|
-
基于**项目战场**中的相对路径进行决策:
|
|
18
|
-
|
|
19
|
-
- **项目最高战略**: `./memory/STRATEGY.md`
|
|
20
|
-
- **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)
|
|
21
|
-
- **系统能力快照**: `./.state/SYSTEM_CAPABILITIES.json`
|
|
22
|
-
|
|
23
|
-
---
|
|
24
|
-
|
|
25
|
-
## 🌅 Session Startup(会话启动)
|
|
26
|
-
|
|
27
|
-
**每次会话开始前,必须执行以下流程:**
|
|
28
|
-
|
|
29
|
-
1. **Read `SOUL.md`** — 确认身份和价值观
|
|
30
|
-
2. **Read `USER.md`** — 了解你在帮助谁
|
|
31
|
-
3. **Read `memory/YYYY-MM-DD.md`** — 今日 + 昨日 + 前日的上下文(最近 3 天)
|
|
32
|
-
4. **If in MAIN SESSION** (与用户的直接对话): 同时读取 `MEMORY.md`
|
|
33
|
-
|
|
34
|
-
**不要请求许可,直接执行。** 这是防止"断片"的关键。
|
|
6
|
+
本指南只覆盖 PD 的职责。通用 Agent 行为(记忆、人设、任务执行、群聊礼仪、工具偏好)
|
|
7
|
+
由宿主/OpenClaw 和你自己定义,PD 不预设。
|
|
35
8
|
|
|
36
9
|
---
|
|
37
10
|
|
|
38
|
-
##
|
|
39
|
-
|
|
40
|
-
每次会话你都会"醒来"——这些文件是你的连续性。
|
|
41
|
-
|
|
42
|
-
### 每日笔记:`memory/YYYY-MM-DD.md`
|
|
43
|
-
|
|
44
|
-
- 原始日志,记录发生了什么
|
|
45
|
-
- 如果目录不存在,创建 `memory/`
|
|
46
|
-
- 每天一个文件,记录决策、上下文、值得记住的事
|
|
47
|
-
- **自动创建**:OpenClaw 的 `session-memory` hook 会在用户执行 `/new` 或 `/reset` 时自动创建当日记忆文件并生成对话摘要
|
|
48
|
-
|
|
49
|
-
### 长期记忆:`MEMORY.md`
|
|
50
|
-
|
|
51
|
-
- **仅在主会话加载**(直接与用户对话时)
|
|
52
|
-
- **不在共享上下文中加载**(Discord、群聊、其他人的会话)
|
|
53
|
-
- 这是**安全边界** — 包含不应泄露给陌生人的私人上下文
|
|
54
|
-
- 你可以自由读取、编辑、更新 `MEMORY.md`
|
|
55
|
-
- 写入重要事件、思考、决策、教训
|
|
56
|
-
- 这是你的**精炼记忆**——精华而非原始日志
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
## 💓 Heartbeats(心跳巡检)
|
|
60
|
-
|
|
61
|
-
当收到心跳轮询时,不要每次只回复 `HEARTBEAT_OK`。利用心跳做有意义的事!
|
|
62
|
-
|
|
63
|
-
### 心跳时应该检查(轮流执行):
|
|
64
|
-
|
|
65
|
-
- **痛觉与进化**: 检查 Runtime V2 诊断(`pd pain record`、candidate、ledger)和 `EVOLUTION_QUEUE.json`;`.pain_flag` 仅为 legacy compatibility
|
|
66
|
-
- **战略对齐**: 对比 `CURRENT_FOCUS.md`,确保未偏离重点
|
|
67
|
-
- **环境健康**: 检查工具链状态、项目根目录整洁度
|
|
68
|
-
|
|
69
|
-
### 心跳状态追踪
|
|
70
|
-
|
|
71
|
-
在 `memory/heartbeat-state.json` 中追踪检查:
|
|
72
|
-
|
|
73
|
-
```json
|
|
74
|
-
{
|
|
75
|
-
"lastChecks": {
|
|
76
|
-
"pain": 1703275200,
|
|
77
|
-
"strategy": 1703260800,
|
|
78
|
-
"grooming": null
|
|
79
|
-
}
|
|
80
|
-
}
|
|
81
|
-
```
|
|
82
|
-
|
|
83
|
-
### 何时主动联系:
|
|
84
|
-
|
|
85
|
-
- 发现重要痛觉信号
|
|
86
|
-
- 战略偏离需要确认
|
|
87
|
-
- 项目环境需要清理
|
|
88
|
-
|
|
89
|
-
### 何时保持沉默(HEARTBEAT_OK):
|
|
11
|
+
## PD 拥有什么
|
|
90
12
|
|
|
91
|
-
-
|
|
92
|
-
-
|
|
93
|
-
-
|
|
94
|
-
-
|
|
13
|
+
- Owner 认为值得改变的行为证据;
|
|
14
|
+
- 可审阅的原则提案;
|
|
15
|
+
- Owner 的批准、拒绝、通道选择、回滚与归档决策;
|
|
16
|
+
- 可逆的激活与后续行为变化的观察。
|
|
95
17
|
|
|
96
|
-
|
|
18
|
+
## PD 不拥有什么
|
|
97
19
|
|
|
98
|
-
|
|
20
|
+
- 通用记忆或记忆维护;
|
|
21
|
+
- Agent 人设或身份;
|
|
22
|
+
- 战略管理或战略对齐;
|
|
23
|
+
- 任务编排或任务衍生;
|
|
24
|
+
- 后台治理(定时任务、超出 Owner 审阅的心跳、环境健康);
|
|
25
|
+
- 用户画像。
|
|
99
26
|
|
|
100
|
-
|
|
101
|
-
2. 识别值得长期保留的重要事件、教训、洞察
|
|
102
|
-
3. 更新 `MEMORY.md`,提炼学习成果
|
|
103
|
-
4. 清理 `MEMORY.md` 中不再相关的过时信息
|
|
104
|
-
|
|
105
|
-
**目标**:有帮助但不烦人。每天检查几次,做有用的后台工作,但尊重安静时间。
|
|
27
|
+
以上属于宿主/OpenClaw 或你本人。
|
|
106
28
|
|
|
107
29
|
---
|
|
108
30
|
|
|
109
|
-
##
|
|
110
|
-
|
|
111
|
-
你有权限访问用户的东西,但这不意味着你应该**分享**他们的东西。
|
|
112
|
-
|
|
113
|
-
在群组中,你是参与者——不是他们的代言人,不是他们的代理。说话前先思考。
|
|
114
|
-
|
|
115
|
-
### 何时回应:
|
|
116
|
-
|
|
117
|
-
- 被直接提及或提问
|
|
118
|
-
- 你能提供真正的价值(信息、洞察、帮助)
|
|
119
|
-
- 有恰到好处的幽默
|
|
120
|
-
- 纠正重要的错误信息
|
|
121
|
-
- 被要求总结
|
|
122
|
-
|
|
123
|
-
### 何时保持沉默(HEARTBEAT_OK):
|
|
124
|
-
|
|
125
|
-
- 人类之间的闲聊
|
|
126
|
-
- 问题已有人回答
|
|
127
|
-
- 你的回复只是"嗯"或"好的"
|
|
128
|
-
- 对话流畅进行中
|
|
129
|
-
- 发言会打断氛围
|
|
130
|
-
|
|
131
|
-
**人类规则**:群聊中的人类不会回复每一条消息。你也不应该。质量 > 数量。
|
|
132
|
-
|
|
133
|
-
### 😊 像人类一样使用 Reactions!
|
|
134
|
-
|
|
135
|
-
在支持 reactions 的平台(Discord、Slack)上,自然地使用 emoji reactions:
|
|
136
|
-
|
|
137
|
-
- 表示欣赏但无需回复(👍, ❤️, 🙌)
|
|
138
|
-
- 觉得好笑(😂, 💀)
|
|
139
|
-
- 觉得有趣或发人深省(🤔, 💡)
|
|
140
|
-
- 想确认但不打断流程
|
|
141
|
-
- 简单的是/否或批准情况(✅, 👀)
|
|
142
|
-
|
|
143
|
-
**不要过度**:每条消息最多一个 reaction。
|
|
144
|
-
|
|
145
|
-
---
|
|
31
|
+
## PD 工作流
|
|
146
32
|
|
|
147
|
-
|
|
148
|
-
|
|
149
|
-
|
|
150
|
-
|
|
151
|
-
- **WhatsApp**: 不支持标题 — 使用 **粗体** 或大写强调
|
|
152
|
-
|
|
153
|
-
---
|
|
154
|
-
|
|
155
|
-
## 🚦 编排者身份 (Orchestrator Mode)
|
|
33
|
+
```text
|
|
34
|
+
行为证据 -> 诊断 -> 原则提案 -> Owner 审阅
|
|
35
|
+
-> 可逆激活 -> 后续可观察的行为变化
|
|
36
|
+
```
|
|
156
37
|
|
|
157
|
-
|
|
38
|
+
- **记录证据**:`pd pain record`(对 Owner 相关的行为证据手动触发)。
|
|
39
|
+
- **审阅提案**:`pd candidate list` 或 `pd console open`——提案等待你的决策(approve / reject / defer / rollback)。
|
|
40
|
+
- **激活**:支持 `prompt`、`code_tool_hook` / RuleHost、`defer_archive` 三种结果。
|
|
41
|
+
- **回滚**:任何已激活原则都可回滚,见 `pd console` 的 Owner 可见控制。
|
|
158
42
|
|
|
159
|
-
|
|
160
|
-
- **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
|
|
43
|
+
## PD 数据位置
|
|
161
44
|
|
|
162
|
-
|
|
45
|
+
- `.principles/` — 原则存储(PRINCIPLES.md、profile、思维模型)。
|
|
46
|
+
- `.state/` — PD 运行时状态(如 pain flag,仅 legacy 兼容)。
|
|
163
47
|
|
|
164
|
-
|
|
165
|
-
- 这是行为建议,不是内置门禁 — PD 默认不强制"先计划后执行"
|
|
166
|
-
- 若 owner 批准的 RuleHost 规则强制了计划行为,该规则会自动生效
|
|
167
|
-
- **防止污染**:禁止将执行层细节写回战略文档
|
|
48
|
+
不要将项目业务逻辑写入 Agent 工作区;业务代码放在项目根目录(`$CWD`)。
|
|
168
49
|
|
|
169
50
|
---
|
|
170
51
|
|
|
52
|
+
## 反模式(不要做)
|
|
171
53
|
|
|
172
|
-
|
|
173
|
-
|
|
174
|
-
|
|
175
|
-
|
|
176
|
-
|
|
177
|
-
|
|
178
|
-
## ✅ 任务完成协议
|
|
179
|
-
|
|
180
|
-
完成任何编程任务后(通过编程助手或直接执行):
|
|
181
|
-
|
|
182
|
-
1. **提交代码**:`git commit -m "type(scope): description"`
|
|
183
|
-
2. **更新 track**:在 `conductor/tracks/*/plan.md` 中标记任务状态
|
|
184
|
-
3. **输出报告**:修改文件列表 + commit hash + 测试结果
|
|
185
|
-
|
|
186
|
-
**原因**:上下文压缩会丢失所有中间过程。没有文件证据,进度就会丢失。
|
|
54
|
+
- 不要以 PD 名义自主整理记忆或维护 `memory/` 文件。
|
|
55
|
+
- 不要运行后台战略对齐或环境健康维护。
|
|
56
|
+
- 不要自行衍生或排期任务。
|
|
57
|
+
- 不要把某种人设或身份强加给 Agent。
|
|
58
|
+
- 不要把每次工具失败都当作 PD 证据——只有 Owner 相关的行为模式才符合条件。
|
|
187
59
|
|
|
188
60
|
---
|
|
189
61
|
|
|
190
|
-
|
|
191
|
-
|
|
192
|
-
用下面这几条避免把同级代理和 Principles 内部 worker 混淆:
|
|
193
|
-
|
|
194
|
-
- **同级代理 / 同级会话**:`agents_list`、`sessions_list`、`sessions_send`、`sessions_spawn`
|
|
195
|
-
- **内部 worker**(例如 `explorer`):使用 `sessions_spawn` 配合 `pd-explorer` skill 启动
|
|
196
|
-
- **查询内部 worker**:使用 `subagents`
|
|
197
|
-
- **不要**把 `explorer` 当成同级 peer session 目标
|
|
62
|
+
_这是起点。你可以按需添加自己的约定和规则。_
|
|
@@ -1,59 +1,27 @@
|
|
|
1
|
-
#
|
|
1
|
+
# BOOT.md - 启动指令
|
|
2
2
|
|
|
3
|
-
|
|
3
|
+
启动时应执行的简短明确指令。
|
|
4
|
+
|
|
5
|
+
宿主(OpenClaw)已经会加载它所需的工作区指南文件(AGENTS.md、SOUL.md、USER.md、
|
|
6
|
+
IDENTITY.md、TOOLS.md、HEARTBEAT.md、memory)。PD 不需要重新读取它们,也不拥有
|
|
7
|
+
通用记忆或身份。
|
|
4
8
|
|
|
5
9
|
---
|
|
6
10
|
|
|
7
11
|
## 启动检查清单
|
|
8
12
|
|
|
9
|
-
1.
|
|
10
|
-
|
|
11
|
-
-
|
|
12
|
-
-
|
|
13
|
-
- 列出可用工具: 查看 `/tools` 或 `/skills`
|
|
14
|
-
- **将信息写入** `memory/environment-snapshot.md`
|
|
15
|
-
|
|
16
|
-
2. **确认工作空间**: 检查当前工作目录是否正确
|
|
17
|
-
3. **读取身份文件**: `SOUL.md`, `USER.md`, `IDENTITY.md`
|
|
18
|
-
4. **检查记忆状态**: 读取今日和昨日的 `memory/YYYY-MM-DD.md`
|
|
19
|
-
5. **检查 Runtime V2 痛苦诊断**: 使用 `pd candidate list` / ledger state;`.state/.pain_flag` 仅为 legacy compatibility
|
|
13
|
+
1. **确认工作空间**:检查当前工作目录是否正确。
|
|
14
|
+
2. **PD 审阅队列**:运行 `pd candidate list`,检查是否有等待 Owner 审阅的原则提案。
|
|
15
|
+
- 如有待审提案,**一次性**给出精简摘要 + 审阅决策选项(approve / reject / defer / rollback)。
|
|
16
|
+
- 否则静默继续。
|
|
20
17
|
|
|
21
18
|
---
|
|
22
19
|
|
|
23
|
-
##
|
|
24
|
-
|
|
25
|
-
创建 `memory/environment-snapshot.md`:
|
|
26
|
-
|
|
27
|
-
```markdown
|
|
28
|
-
# 环境快照
|
|
29
|
-
|
|
30
|
-
> 最后更新: [日期时间]
|
|
31
|
-
|
|
32
|
-
## 系统信息
|
|
33
|
-
|
|
34
|
-
- **操作系统**: [Windows/Linux/macOS]
|
|
35
|
-
- **工作目录**: [完整路径]
|
|
36
|
-
- **Git 仓库**: [是/否] - [分支名]
|
|
37
|
-
- **Shell**: [bash/zsh/powershell等]
|
|
20
|
+
## 边界
|
|
38
21
|
|
|
39
|
-
|
|
40
|
-
|
|
41
|
-
-
|
|
42
|
-
- [技能列表]
|
|
43
|
-
|
|
44
|
-
## 目录结构
|
|
45
|
-
|
|
46
|
-
```
|
|
47
|
-
[简要列出关键目录]
|
|
48
|
-
```
|
|
49
|
-
|
|
50
|
-
## 环境变量
|
|
51
|
-
|
|
52
|
-
[重要的环境变量,如 PATH 等]
|
|
53
|
-
|
|
54
|
-
---
|
|
55
|
-
_此文件应在每次启动时更新_
|
|
56
|
-
```
|
|
22
|
+
- **不要**在启动时写入环境快照或运行时状态文件——环境发现是宿主/OpenClaw 自身能力,
|
|
23
|
+
PD 不拥有通用记忆或环境持久化。
|
|
24
|
+
- **不要**在 PD 启动流程中读取或管理 `memory/` 文件——那是宿主/OpenClaw 的职责。
|
|
57
25
|
|
|
58
26
|
---
|
|
59
27
|
|
|
@@ -1,98 +1,36 @@
|
|
|
1
|
-
# 💓 HEARTBEAT.md -
|
|
1
|
+
# 💓 HEARTBEAT.md - 心跳巡检(Owner 治理最小版)
|
|
2
2
|
|
|
3
|
-
|
|
4
|
-
|
|
5
|
-
|
|
3
|
+
PD 是 Owner 治理的行为内化层。心跳只做一件事:**浮现需要 Owner 审阅的原则提案**。
|
|
4
|
+
其余一切(环境检查、记忆整理、任务衍生、后台 Cron 维护)要么是宿主/OpenClaw 自身能力,
|
|
5
|
+
要么超出 PD 的 MVP 边界——PD 不拥有通用记忆、任务编排或自主决策。
|
|
6
6
|
|
|
7
7
|
---
|
|
8
8
|
|
|
9
|
-
##
|
|
9
|
+
## 检查:Owner 审阅队列
|
|
10
10
|
|
|
11
|
-
- [ ]
|
|
12
|
-
|
|
13
|
-
- [ ] **`memory/logs/SYSTEM.log`** — 最近是否有未解决的高优先级问题?
|
|
11
|
+
- [ ] 运行 `pd candidate list`(或 `pd console open`)查看 ledger 中是否有
|
|
12
|
+
等待 Owner 审阅的原则提案。
|
|
14
13
|
|
|
15
|
-
|
|
14
|
+
**行动**:如有待审提案,**一次性**给出精简摘要 + 审阅决策选项(approve / reject /
|
|
15
|
+
defer / rollback)。否则保持沉默。
|
|
16
16
|
|
|
17
17
|
---
|
|
18
18
|
|
|
19
|
-
##
|
|
20
|
-
|
|
21
|
-
### 检查与压缩
|
|
22
|
-
|
|
23
|
-
- [ ] **检查文件大小**:目标 < 40 行
|
|
24
|
-
- [ ] **如 > 40 行**:执行压缩流程(见下方)
|
|
25
|
-
|
|
26
|
-
### 压缩流程
|
|
19
|
+
## 保持沉默(HEARTBEAT_OK)
|
|
27
20
|
|
|
28
|
-
|
|
29
|
-
2. **递增版本号**:修改 `**版本**: v{N+1}`
|
|
30
|
-
3. **更新日期**:修改 `**更新**: {today}`
|
|
31
|
-
4. **压缩内容**:
|
|
32
|
-
- 移除已完成的 P0/P1 任务(保留 P2)
|
|
33
|
-
- 将重要里程碑移到 `MEMORY.md`
|
|
34
|
-
- 保留:状态快照、当前任务、下一步
|
|
35
|
-
5. **清理历史**:保留最近 10 个历史版本
|
|
21
|
+
以下情况回复 `HEARTBEAT_OK`:
|
|
36
22
|
|
|
37
|
-
|
|
23
|
+
- 没有待审的 Owner 审阅项;
|
|
24
|
+
- 提案最近已浮现且 Owner 尚未处理;
|
|
25
|
+
- 深夜(23:00-08:00)且非紧急事项。
|
|
38
26
|
|
|
39
|
-
|
|
40
|
-
# 🎯 CURRENT_FOCUS
|
|
41
|
-
|
|
42
|
-
> **版本**: v{N} | **状态**: {状态} | **更新**: {日期}
|
|
43
|
-
|
|
44
|
-
## 📍 状态快照 ← 必须保留
|
|
45
|
-
## 🔄 当前任务 ← 必须保留
|
|
46
|
-
## ➡️ 下一步 ← 必须保留(summary 模式优先提取)
|
|
47
|
-
## 📎 参考 ← 可选
|
|
48
|
-
```
|
|
27
|
+
无事发生是正常状态。不要为了"保持有用"而发明工作、衍生任务、整理记忆或主动打扰 Owner。
|
|
49
28
|
|
|
50
29
|
---
|
|
51
30
|
|
|
52
|
-
##
|
|
53
|
-
|
|
54
|
-
> **为什么在心跳里?** 大模型健忘问题严重,需要频繁提醒。心跳有主会话上下文,能感知当前工作内容。
|
|
55
|
-
|
|
56
|
-
### 每次心跳执行:
|
|
57
|
-
|
|
58
|
-
- [ ] **回顾今日笔记** `memory/YYYY-MM-DD.md` — 今天学了什么?做了什么决策?
|
|
59
|
-
- [ ] **检查关键记忆** `MEMORY.md` — 是否有需要更新的内容?
|
|
60
|
-
|
|
61
|
-
### 当 memory 文件积累较多时:
|
|
62
|
-
|
|
63
|
-
- [ ] **提炼精华**: 将每日笔记中的重要内容提炼到 `MEMORY.md`
|
|
64
|
-
- [ ] **清理过时信息**: 移除 `MEMORY.md` 中不再相关的内容
|
|
65
|
-
|
|
66
|
-
**原则**:
|
|
67
|
-
- 每日文件是原始笔记,`MEMORY.md` 是精炼的智慧
|
|
68
|
-
- **不要让重要决策和教训被遗忘**
|
|
69
|
-
- 如果用户分享了偏好或重要信息,立即记录
|
|
70
|
-
|
|
71
|
-
---
|
|
72
|
-
|
|
73
|
-
## ⏰ 保持沉默(HEARTBEAT_OK):
|
|
74
|
-
|
|
75
|
-
- 深夜(23:00-08:00)除非紧急 / 用户在忙 / 无新情况 / < 30 分钟 / 正常
|
|
76
|
-
|
|
77
|
-
---
|
|
78
|
-
|
|
79
|
-
## 🚨 主动联系:
|
|
80
|
-
|
|
81
|
-
- 痛觉需处理 / 重要发现 / 进化任务需执行 / 记忆需要更新
|
|
82
|
-
|
|
83
|
-
---
|
|
84
|
-
|
|
85
|
-
## 📋 任务队列检查(WEEK_TASKS.json)
|
|
86
|
-
|
|
87
|
-
- [ ] **文件存在?** `okr/WEEK_TASKS.json` — 若不存在,从模板创建
|
|
88
|
-
- [ ] **队列为空?** 无 `pending` 或 `in_progress` 任务 → 查看 RECOVERY_PROTOCOL.md
|
|
89
|
-
- [ ] **任务卡住?** `in_progress` 任务的 `startedAt` > 2 小时 → 与用户确认
|
|
90
|
-
- [ ] **周切换?** 检查 `week` 字段是否为当前周 → 归档旧任务
|
|
91
|
-
|
|
92
|
-
**行动**:
|
|
93
|
-
- 空队列 + 无指令 → 按 RECOVERY_PROTOCOL.md 自主推导
|
|
94
|
-
- 任务卡住 → 询问用户:"任务 [id] 似乎卡住了。继续还是归档?"
|
|
95
|
-
|
|
96
|
-
---
|
|
31
|
+
## 原则
|
|
97
32
|
|
|
98
|
-
|
|
33
|
+
- **必要**:只有 Owner 审阅项够格。
|
|
34
|
+
- **低噪音**:浮现一次,然后等待。
|
|
35
|
+
- **显式授权**:无后台任务、无自衍生任务、无自动记忆维护。
|
|
36
|
+
- **直接连接 PD 治理**:行为证据 → 诊断 → 提案 → Owner 审阅 → 可逆激活。
|
|
@@ -1,8 +1,15 @@
|
|
|
1
|
-
#
|
|
2
|
-
|
|
3
|
-
|
|
4
|
-
|
|
5
|
-
|
|
6
|
-
|
|
7
|
-
|
|
8
|
-
|
|
1
|
+
# IDENTITY.md - 身份占位文件
|
|
2
|
+
|
|
3
|
+
本文件供宿主/OpenClaw 用于 Agent 身份(名称、风格、emoji)。
|
|
4
|
+
|
|
5
|
+
Principles Disciple **不预设也不强制**任何人设。请填入你自己的 Agent 身份,
|
|
6
|
+
或保留此占位文件。
|
|
7
|
+
|
|
8
|
+
<!-- 替换为你自己的身份,例如:
|
|
9
|
+
|
|
10
|
+
- **Name:** <你的 Agent 名称>
|
|
11
|
+
- **Role:** <这个 Agent 做什么>
|
|
12
|
+
- **Vibe:** <语气与风格>
|
|
13
|
+
- **Emoji:** <一个 emoji>
|
|
14
|
+
|
|
15
|
+
-->
|
|
@@ -1,7 +1,7 @@
|
|
|
1
|
-
|
|
2
|
-
|
|
3
|
-
|
|
4
|
-
|
|
5
|
-
-
|
|
6
|
-
|
|
7
|
-
|
|
1
|
+
# Principles
|
|
2
|
+
|
|
3
|
+
Principles Disciple 默认不会替 Owner 预装其个人行为原则。
|
|
4
|
+
|
|
5
|
+
T-01~T-10 内置核心公理由 PD Runtime 提供。
|
|
6
|
+
|
|
7
|
+
由 Owner 审核产生的原则由 PD 原则仓库与运行时生命周期负责管理。
|