openclaw-node-harness 2.0.0

package/skills/byterover/.clawhub/origin.json

@@ -0,0 +1,7 @@
+{
+  "version": 1,
+  "registry": "https://clawhub.ai",
+  "slug": "byterover",
+  "installedVersion": "1.2.2",
+  "installedAt": 1770236813771
+}

package/skills/byterover/SKILL.md

@@ -0,0 +1,105 @@
+---
+name: byterover
+description: "Manages project knowledge using ByteRover context tree. Query to retrieve patterns/conventions, curate to persist decisions and insights. Use when you need cross-session project knowledge or want to store learned patterns."
+triggers:
+  - "byterover"
+  - "brv query"
+  - "project knowledge lookup"
+  - "persist project context"
+  - "curate knowledge"
+negative_triggers:
+  - "clawvault"
+  - "memorylayer"
+  - "local memory files"
+  - "MEMORY.md"
+metadata:
+  author: ByteRover Inc. (https://byterover.dev/)
+  version: "1.2.1"
+---
+
+# ByteRover Context Tree
+
+A project-level knowledge repository that persists across sessions. Use it to avoid re-discovering patterns, conventions, and decisions.
+
+## Why Use ByteRover
+
+- **Query before working**: Get existing knowledge about patterns, conventions, and past decisions before implementing
+- **Curate after learning**: Capture insights, decisions, and bug fixes so future sessions start informed
+
+## Quick Reference
+
+| Command | When | Example |
+|---------|------|---------|
+| `brv query "question"` | Before starting work | `brv query "How is auth implemented?"` |
+| `brv curate "context" -f file` | After completing work | `brv curate "JWT 24h expiry" -f auth.ts` |
+| `brv status` | To check prerequisites | `brv status` |
+
+## When to Use
+
+**Query** when you need to understand something:
+- "How does X work in this codebase?"
+- "What patterns exist for Y?"
+- "Are there conventions for Z?"
+
+**Curate** when you learned or created something valuable:
+- Implemented a feature using specific patterns
+- Fixed a bug and found root cause
+- Made an architecture decision
+
+## Curate Quality
+
+Context must be **specific** and **actionable**:
+
+```bash
+# Good - specific, explains where and why
+brv curate "Auth uses JWT 24h expiry, tokens in httpOnly cookies" -f src/auth.ts
+
+# Bad - too vague
+brv curate "Fixed auth"
+```
+
+**Note:** Context argument must come before `-f` flags. Max 5 files.
+
+## Best Practices
+
+1. **Break down large contexts** - Run multiple `brv curate` commands for complex topics rather than one massive context. Smaller chunks are easier to retrieve and update.
+
+2. **Let ByteRover read files** - Don't read files yourself before curating. Use `-f` flags to let ByteRover read them directly:
+   ```bash
+   # Good - ByteRover reads the files
+   brv curate "Auth implementation details" -f src/auth.ts -f src/middleware/jwt.ts
+
+   # Wasteful - reading files twice
+   # [agent reads files] then brv curate "..." -f same-files
+   ```
+
+3. **Be specific in queries** - Queries block your workflow. Use precise questions to get faster, more relevant results:
+   ```bash
+   # Good - specific
+   brv query "What validation library is used for API request schemas?"
+
+   # Bad - vague, slow
+   brv query "How is validation done?"
+   ```
+
+4. **Signal outdated context** - When curating updates that replace existing knowledge, explicitly tell ByteRover to clean up:
+   ```bash
+   brv curate "OUTDATED: Previous auth used sessions. NEW: Now uses JWT with refresh tokens. Clean up old session-based auth context." -f src/auth.ts
+   ```
+
+5. **Specify structure expectations** - Guide ByteRover on how to organize the knowledge:
+   ```bash
+   # Specify topics/domains
+   brv curate "Create separate topics for: 1) JWT validation, 2) refresh token flow, 3) logout handling" -f src/auth.ts
+
+   # Specify detail level
+   brv curate "Document the error handling patterns in detail (at least 30 lines covering all error types)" -f src/errors/
+   ```
+
+## Prerequisites
+
+Run `brv status` first. If errors occur, the agent cannot fix them—instruct the user to take action in their brv terminal. See [TROUBLESHOOTING.md](TROUBLESHOOTING.md) for details.
+
+---
+
+**See also:** [WORKFLOWS.md](WORKFLOWS.md) for detailed patterns and examples, [TROUBLESHOOTING.md](TROUBLESHOOTING.md) for error handling

package/skills/byterover/TROUBLESHOOTING.md

@@ -0,0 +1,50 @@
+# ByteRover Troubleshooting
+
+## Quick Diagnosis
+
+```bash
+brv status
+```
+
+## User Action Required
+
+These errors require user intervention (agent cannot fix):
+
+| Error | User Action |
+|-------|-------------|
+| "No ByteRover instance is running" | Start `brv` in separate terminal |
+| "Not authenticated" | Run `/login` in brv REPL |
+| "Project not initialized" | Run `/init` in brv REPL |
+| "Connection failed" | Restart `brv` (Ctrl+C, then `brv`) |
+
+**Template response:**
+> Please [action] in your brv terminal, then I'll retry the command.
+
+## Agent-Fixable Errors
+
+| Error | Fix |
+|-------|-----|
+| "Context argument required" | Add text before `-f`: `brv curate "text" -f file` |
+| "Maximum 5 files allowed" | Reduce to 5 or fewer `-f` flags |
+| "File not found" | Verify path with `ls`, use relative paths from project root |
+| "No relevant context found" | Try different query phrasing, or curate knowledge first |
+
+## Architecture
+
+ByteRover uses client-server architecture:
+- User runs `brv` to start server (interactive REPL)
+- Agent commands (`query`, `curate`, `status`) connect to server
+- Server must be running for commands to work
+
+## Exit Codes
+
+| Code | Meaning |
+|------|---------|
+| 0 | Success |
+| 1 | General error |
+| 2 | Connection error |
+
+## Getting Help
+
+- Email: support@byterover.dev
+- Discord: https://discord.com/invite/UMRrpNjh5W

package/skills/byterover/WORKFLOWS.md

@@ -0,0 +1,229 @@
+# ByteRover Workflows
+
+## Pattern 1: Research Before Implementation
+
+Use when starting new features or working in unfamiliar areas.
+
+**Workflow:**
+```
+Query existing knowledge → Implement following patterns → Curate new decisions
+```
+
+**Example: Adding a new API endpoint**
+
+```bash
+# 1. Query for existing patterns
+# Be SPECIFIC to get faster, more relevant results
+brv query "What middleware chain is used for authenticated API endpoints?"
+brv query "What is the standard error response format for API routes?"
+
+# 2. Implement following discovered patterns
+# ... write code ...
+
+# 3. Curate decisions made
+# Don't read the file first - let ByteRover read it via -f flag
+brv curate "Added /api/users/profile endpoint. Uses authMiddleware for JWT validation, returns UserProfileDTO. Error handling follows ApiError pattern" -f src/api/routes/users.ts
+```
+
+**Example: Implementing a new React component**
+
+```bash
+# 1. Query patterns - specific questions save time
+brv query "What data fetching hook pattern is used in list components?"
+brv query "Are CSS modules or styled-components used for component styling?"
+
+# 2. Implement
+# ... write component ...
+
+# 3. Curate with file reference - ByteRover reads the file
+brv curate "Created UserAvatar component. Uses React Query for data fetching, CSS modules for styling. Follows existing pattern in src/components/common/" -f src/components/UserAvatar.tsx
+```
+
+## Pattern 2: Debug and Document
+
+Use when fixing bugs to capture learnings and prevent recurrence.
+
+**Workflow:**
+```
+Query known issues → Debug and fix → Curate root cause and solution
+```
+
+**Example: Fixing a race condition**
+
+```bash
+# 1. Query for context - be specific about the symptom
+brv query "Are there known race condition patterns or AbortController usage in data fetching hooks?"
+
+# 2. Debug and fix
+# ... investigate, find root cause, implement fix ...
+
+# 3. Curate the learning with structured format
+brv curate "Bug: stale data after rapid navigation. Cause: useEffect cleanup not cancelling requests. Fix: AbortController with cleanup. Pattern: always abort fetch in useEffect cleanup" -f src/hooks/useUserData.ts
+```
+
+**Example: Fixing an authentication issue**
+
+```bash
+# 1. Specific query about the problem area
+brv query "How are credentials and cookies configured in the API client fetch calls?"
+
+# 2. Fix
+# ... debug and fix ...
+
+# 3. Curate with emphasis on the gotcha
+brv curate "Bug: unexpected logouts. Cause: fetch missing credentials option. Fix: added 'credentials: include' to fetch config. IMPORTANT: all API calls must include credentials for cookie-based auth" -f src/lib/api-client.ts
+```
+
+## Pattern 3: Multi-File Changes
+
+Use when changes span multiple files to maintain context.
+
+**Example: Adding a new feature across layers**
+
+```bash
+# 1. Query architecture - specific about the layers
+brv query "How are full-stack features organized across API routes, service layer, and React components?"
+
+# 2. Implement across files
+# ... write code ...
+
+# 3. Curate with multiple files (max 5)
+# Let ByteRover read all files - don't read them yourself first
+# Specify the topics you want created
+brv curate "Added user notifications feature. Create separate topics for: 1) API endpoint structure, 2) NotificationService business logic, 3) useNotifications React hook pattern, 4) NotificationBell UI component" -f src/api/notifications.ts -f src/services/NotificationService.ts -f src/hooks/useNotifications.ts -f src/components/NotificationBell.tsx
+```
+
+**Example: Database schema change with migrations**
+
+```bash
+# 1. Query migration patterns
+brv query "What is the migration naming convention and how are schema changes tested?"
+
+# 2. Implement
+# ... create migration, update models, update queries ...
+
+# 3. Multiple curates for different concerns - break down large context
+brv curate "Added user_preferences table with JSONB settings column. Migration: 20240115_add_user_preferences" -f migrations/20240115_add_user_preferences.ts
+
+brv curate "UserPreferences model with type-safe JSONB access. Includes validation for theme, notifications, locale settings" -f src/models/UserPreferences.ts
+
+brv curate "Updated UserService to load/save preferences. Uses transaction for atomic updates with user record" -f src/services/UserService.ts
+```
+
+## Pattern 4: Updating Existing Knowledge
+
+Use when changes make previous context outdated.
+
+**Example: Refactoring authentication system**
+
+```bash
+# 1. Query current documented state
+brv query "What is currently documented about authentication implementation and token handling?"
+
+# 2. Implement refactor
+# ... refactor code ...
+
+# 3. Curate with explicit cleanup signal
+# Tell ByteRover to clean up outdated context
+brv curate "OUTDATED: Previous auth used session cookies stored in Redis. NEW: Migrated to JWT with refresh tokens. Access token in memory (15min), refresh token in httpOnly cookie (7d). Remove/update any session-based auth context in the tree" -f src/auth/jwt.ts -f src/auth/refresh.ts
+```
+
+**Example: Replacing a library**
+
+```bash
+# 1. Query what's documented about the old library
+brv query "What patterns are documented for moment.js date handling?"
+
+# 2. Implement replacement
+# ... replace library usage ...
+
+# 3. Curate the replacement with cleanup signal
+brv curate "REPLACED: Removed moment.js (bloated, mutable). Now using date-fns (tree-shakeable, immutable). Clean up any moment.js context. New patterns: format(date, 'yyyy-MM-dd'), parseISO(string), differenceInDays(date1, date2)" -f src/utils/dates.ts
+```
+
+**Example: API versioning change**
+
+```bash
+# Curate with explicit version context
+brv curate "OUTDATED: API v1 endpoints deprecated. NEW: All endpoints now v2 with breaking changes. v2 uses camelCase response keys, pagination via cursor (not offset), errors include requestId. Update any v1 API context" -f src/api/v2/routes.ts
+```
+
+## Pattern 5: Comprehensive Documentation
+
+Use when documenting complex systems that need detailed breakdown.
+
+**Example: Documenting a payment module**
+
+```bash
+# Break into multiple curates - don't try to capture everything in one
+# Specify structure and detail level for each topic
+
+# Overview first
+brv curate "Payment module overview: Stripe integration with webhooks for subscription management. Create 4 separate detailed topics covering the full payment lifecycle" -f src/payments/
+
+# Topic 1 - detailed with line count guidance
+brv curate "Topic: Payment checkout flow. Cover: 1) cart validation, 2) createPaymentIntent call, 3) client-side confirmation, 4) success/failure handling. Include error scenarios. At least 50 lines of detailed documentation" -f src/payments/checkout.ts -f src/payments/intent.ts
+
+# Topic 2 - webhook handling
+brv curate "Topic: Stripe webhook handling. Cover: 1) signature verification with STRIPE_WEBHOOK_SECRET, 2) idempotency with processed_events table, 3) event types (payment_intent.succeeded, payment_intent.failed, customer.subscription.*). At least 40 lines" -f src/payments/webhooks.ts
+
+# Topic 3 - subscription management
+brv curate "Topic: Subscription lifecycle. Cover: trial period handling, upgrade/downgrade proration, cancellation with grace period, reactivation flow" -f src/payments/subscriptions.ts
+
+# Topic 4 - error handling
+brv curate "Topic: Payment error handling. Cover: StripeError types (CardError, InvalidRequestError, APIError), retry logic for transient failures, user-facing error messages mapping" -f src/payments/errors.ts
+```
+
+**Example: Documenting a state management system**
+
+```bash
+# Multiple focused curates instead of one massive one
+
+brv curate "State management overview: Using Zustand with persistence middleware. Create topics for: store structure, async actions, persistence, devtools integration" -f src/store/
+
+brv curate "Topic: Store structure. Separate stores per domain (userStore, cartStore, uiStore). Each store follows pattern: state interface, actions, selectors. No cross-store dependencies" -f src/store/userStore.ts -f src/store/cartStore.ts
+
+brv curate "Topic: Async actions. Pattern: set loading -> try/catch -> set data/error -> clear loading. All API calls go through apiClient. Optimistic updates for cart operations" -f src/store/cartStore.ts -f src/store/actions/
+
+brv curate "Topic: Persistence. userStore persisted to localStorage (excluding sensitive data). cartStore persisted to sessionStorage. Custom serializer excludes functions and timestamps" -f src/store/middleware/persist.ts
+```
+
+## Pattern 6: Exploratory Documentation
+
+Use when you need to document an unfamiliar codebase area.
+
+**Example: Understanding and documenting an existing module**
+
+```bash
+# 1. Query what's already known
+brv query "What is documented about the notification system and real-time updates?"
+
+# 2. Explore the code (agent reads files to understand)
+# ... read and understand the code ...
+
+# 3. Curate in chunks as you understand different parts
+# First curate: high-level architecture
+brv curate "Notification system uses WebSocket for real-time delivery with Redis pub/sub for horizontal scaling. Three components: NotificationService (creation/storage), NotificationGateway (WebSocket), NotificationWorker (background processing)" -f src/notifications/
+
+# Second curate: specific implementation detail
+brv curate "WebSocket authentication: JWT token passed in connection query params, validated on connect, stored in socket.data. Rooms: user_{id} for personal, team_{id} for team broadcasts" -f src/notifications/NotificationGateway.ts
+
+# Third curate: gotchas discovered
+brv curate "Notification gotchas: 1) Must call gateway.joinRoom after auth, 2) Unread count cached in Redis (5min TTL) - call invalidateUnreadCount after marking read, 3) Batch notifications throttled to max 10/second per user" -f src/notifications/NotificationService.ts
+```
+
+## What to Curate
+
+**Do curate:**
+- Architecture decisions: "Chose Redis for sessions because of horizontal scaling"
+- Patterns: "All forms use react-hook-form with zod. Pattern in LoginForm.tsx"
+- Non-obvious conventions: "File uploads go to /tmp first, then S3 after validation"
+- Bug root causes: "Memory leak from event listeners not removed on unmount"
+- Gotchas: "PostgreSQL JSONB queries need explicit casting for arrays"
+- Replacements: "OUTDATED: X, NEW: Y" when refactoring
+
+**Don't curate:**
+- Obvious facts: "Uses TypeScript", "Has a README"
+- Temporary states: "Currently debugging X"
+- Personal preferences: "I prefer tabs"
+- Trivial changes: "Fixed typo in comment"

package/skills/capability-evolver/.clawhub/origin.json

@@ -0,0 +1,7 @@
+{
+  "version": 1,
+  "registry": "https://clawhub.ai",
+  "slug": "capability-evolver",
+  "installedVersion": "1.1.0",
+  "installedAt": 1770213018296
+}

package/skills/capability-evolver/CONTRIBUTING.md

@@ -0,0 +1,11 @@
+## Contributing
+
+Thank you for contributing. Please follow these rules:
+
+- Do not use emoji (except the DNA emoji in documentation if needed).
+- Keep changes small and reviewable.
+- Update related documentation when you change behavior.
+- Run `node index.js` for a quick sanity check.
+
+Submit PRs with clear intent and scope.
+

package/skills/capability-evolver/README.md

@@ -0,0 +1,157 @@
+# 🧬 Capability Evolver
+
+![Capability Evolver Cover](assets/cover.png)
+
+[Chinese Docs](README.zh-CN.md)
+
+**"Evolution is not optional. Adapt or die."**
+
+**Three lines**
+- **What it is**: A protocol-constrained self-evolution engine for AI agents.
+- **Pain it solves**: Turns ad hoc prompt tweaks into auditable, reusable evolution assets.
+- **Use in 30 seconds**: `node index.js` to generate a GEP-guided evolution prompt.
+
+Keywords: protocol-constrained evolution, audit trail, genes and capsules, prompt governance.
+
+## Try It Now (Minimal)
+
+```bash
+node index.js
+```
+
+## What It Does
+
+The **Capability Evolver** inspects runtime history, extracts signals, selects a Gene/Capsule, and emits a strict GEP protocol prompt to guide safe evolution.
+
+## Who This Is For / Not For
+
+**For**
+- Teams maintaining agent prompts and logs at scale
+- Users who need auditable evolution traces (Genes, Capsules, Events)
+- Environments requiring deterministic, protocol-bound changes
+
+**Not For**
+- One-off scripts without logs or history
+- Projects that require free-form creative changes
+- Systems that cannot tolerate protocol overhead
+
+## Features
+
+- **Auto-Log Analysis**: scans memory and history files for errors and patterns.
+- **Self-Repair Guidance**: emits repair-focused directives from signals.
+- **GEP Protocol**: standardized evolution with reusable assets.
+- **One-Command Evolution**: `node index.js` to generate the prompt.
+
+## Typical Use Cases
+
+- Harden a flaky agent loop by enforcing validation before edits
+- Encode recurring fixes as reusable Genes and Capsules
+- Produce auditable evolution events for review or compliance
+
+## Anti-Examples
+
+- Rewriting entire subsystems without signals or constraints
+- Using the protocol as a generic task runner
+- Producing changes without recording EvolutionEvent
+
+## FAQ
+
+**Does this edit code automatically?**
+No. It generates a protocol-bound prompt and assets that guide evolution.
+
+**Do I need to use all GEP assets?**
+No. You can start with default Genes and extend over time.
+
+**Is this safe in production?**
+Use review mode and validation steps. Treat it as a safety-focused evolution tool, not a live patcher.
+
+## Roadmap
+
+- Add a one-minute demo workflow
+- Add a public changelog
+- Add a comparison table vs alternatives
+
+## GEP Protocol (Auditable Evolution)
+
+This repo includes a protocol-constrained prompt mode based on GEP (Genome Evolution Protocol).
+
+- **Structured assets** live in `assets/gep/`:
+  - `assets/gep/genes.json`
+  - `assets/gep/capsules.json`
+  - `assets/gep/events.jsonl`
+- **Selector** logic uses extracted signals to prefer existing Genes/Capsules and emits a JSON selector decision in the prompt.
+- **Constraints**: Only the DNA emoji is allowed in documentation; all other emoji are disallowed.
+
+## Usage
+
+### Standard Run (Automated)
+```bash
+node index.js
+```
+
+### Review Mode (Human-in-the-Loop)
+```bash
+node index.js --review
+```
+
+### Continuous Loop
+```bash
+node index.js --loop
+```
+
+## Public Release
+
+This repository is the public distribution.
+
+- Build public output: `npm run build`
+- Publish public output: `npm run publish:public`
+- Dry run: `DRY_RUN=true npm run publish:public`
+
+Required env vars:
+
+- `PUBLIC_REMOTE` (default: `public`)
+- `PUBLIC_REPO` (e.g. `autogame-17/evolver`)
+ - `PUBLIC_OUT_DIR` (default: `dist-public`)
+ - `PUBLIC_USE_BUILD_OUTPUT` (default: `true`)
+
+Optional env vars:
+
+- `SOURCE_BRANCH` (default: `main`)
+- `PUBLIC_BRANCH` (default: `main`)
+- `RELEASE_TAG` (e.g. `v1.0.41`)
+- `RELEASE_TITLE` (e.g. `v1.0.41 - GEP protocol`)
+- `RELEASE_NOTES` or `RELEASE_NOTES_FILE`
+- `GITHUB_TOKEN` (or `GH_TOKEN` / `GITHUB_PAT`) for GitHub Release creation
+- `RELEASE_SKIP` (`true` to skip creating a GitHub Release; default is to create)
+- `RELEASE_USE_GH` (`true` to use `gh` CLI instead of GitHub API)
+- `PUBLIC_RELEASE_ONLY` (`true` to only create a Release for an existing tag; no publish)
+
+## Versioning (SemVer)
+
+MAJOR.MINOR.PATCH
+
+- MAJOR: incompatible changes
+- MINOR: backward-compatible features
+- PATCH: backward-compatible bug fixes
+
+## Configuration & Decoupling
+
+This skill is designed to be **environment-agnostic**. It uses standard OpenClaw tools by default.
+
+### Local Overrides (Injection)
+You can inject local preferences (e.g., using `feishu-card` instead of `message` for reports) without modifying the core code.
+
+**Method 1: Environment Variables**
+Set `EVOLVE_REPORT_TOOL` in your `.env` file:
+```bash
+EVOLVE_REPORT_TOOL=feishu-card
+```
+
+**Method 2: Dynamic Detection**
+The script automatically detects if compatible local skills (like `skills/feishu-card`) exist in your workspace and upgrades its behavior accordingly.
+
+## License
+
+MIT
+
+

package/skills/capability-evolver/README.zh-CN.md

@@ -0,0 +1,112 @@
+# 🧬 Capability Evolver（能力进化引擎）
+
+[English Docs](README.md)
+
+**“进化不是可选项，而是生存法则。”**
+
+**Capability Evolver** 是一个元技能（Meta-Skill），赋予 OpenClaw 智能体自我反省的能力。它可以扫描自身的运行日志，识别效率低下或报错的地方，并自主编写代码补丁来优化自身性能。
+
+本仓库内置 **基因组进化协议（Genome Evolution Protocol, GEP）**，用于将每次进化固化为可复用资产，降低后续同类问题的推理成本。
+
+## 核心特性
+
+- **自动日志分析**：自动扫描 `.jsonl` 会话日志，寻找错误模式。
+- **自我修复**：检测运行时崩溃并编写修复补丁。
+- **GEP 协议**：标准化进化流程与可复用资产，支持可审计与可共享。
+- **动态集成**：自动检测并使用本地工具（如 `git-sync` 或 `feishu-card`），如果不存在则回退到通用模式，零依赖运行。
+- **持续循环模式**：持续运行的自我修复循环。
+
+## 使用方法
+
+### 标准运行（自动化）
+```bash
+node index.js
+```
+
+### 审查模式（人工介入）
+在应用更改前暂停，等待人工确认。
+```bash
+node index.js --review
+```
+
+### 持续循环（守护进程）
+无限循环运行。适合作为后台服务。
+```bash
+node index.js --loop
+```
+
+## 典型使用场景
+
+- 需要审计与可追踪的提示词演进
+- 团队协作维护 Agent 的长期能力
+- 希望将修复经验固化为可复用资产
+
+## 反例
+
+- 一次性脚本或没有日志的场景
+- 需要完全自由发挥的改动
+- 无法接受协议约束的系统
+
+## GEP 协议（可审计进化）
+
+本仓库内置基于 GEP 的“协议受限提示词模式”，用于把每次进化固化为可复用资产。
+
+- **结构化资产目录**：`assets/gep/`
+  - `assets/gep/genes.json`
+  - `assets/gep/capsules.json`
+  - `assets/gep/events.jsonl`
+- **Selector 选择器**：根据日志提取 signals，优先复用已有 Gene/Capsule，并在提示词中输出可审计的 Selector 决策 JSON。
+- **约束**：除 🧬 外，禁止使用其他 emoji。
+
+## 配置与解耦
+
+本插件能自动适应你的环境。
+
+| 环境变量 | 描述 | 默认值 |
+| :--- | :--- | :--- |
+| `EVOLVE_REPORT_TOOL` |用于报告结果的工具名称（例如 `feishu-card`） | `message` |
+| `MEMORY_DIR` | 记忆文件路径 | `./memory` |
+
+## Public 发布
+
+本仓库为公开发行版本。
+
+- 构建公开产物：`npm run build`
+- 发布公开产物：`npm run publish:public`
+- 演练：`DRY_RUN=true npm run publish:public`
+
+必填环境变量：
+
+- `PUBLIC_REMOTE`（默认：`public`）
+- `PUBLIC_REPO`（例如 `autogame-17/evolver`）
+- `PUBLIC_OUT_DIR`（默认：`dist-public`）
+- `PUBLIC_USE_BUILD_OUTPUT`（默认：`true`）
+
+可选环境变量：
+
+- `SOURCE_BRANCH`（默认：`main`）
+- `PUBLIC_BRANCH`（默认：`main`）
+- `RELEASE_TAG`（例如 `v1.0.41`）
+- `RELEASE_TITLE`（例如 `v1.0.41 - GEP protocol`）
+- `RELEASE_NOTES` 或 `RELEASE_NOTES_FILE`
+- `GITHUB_TOKEN`（或 `GH_TOKEN` / `GITHUB_PAT`，用于创建 GitHub Release）
+- `RELEASE_SKIP`（`true` 则跳过创建 GitHub Release；默认会创建）
+- `RELEASE_USE_GH`（`true` 则使用 `gh` CLI，否则默认走 GitHub API）
+- `PUBLIC_RELEASE_ONLY`（`true` 则仅为已存在的 tag 创建 Release；不发布代码）
+
+## 版本号规则（SemVer）
+
+MAJOR.MINOR.PATCH
+
+• MAJOR（主版本）：有不兼容变更  
+• MINOR（次版本）：向后兼容的新功能  
+• PATCH（修订/补丁）：向后兼容的问题修复
+
+## 安全协议
+
+1.  **单进程锁**：进化引擎禁止生成子进化进程（防止 Fork 炸弹）。
+2.  **稳定性优先**：如果近期错误率较高，强制进入 **修复模式**，暂停创新功能。
+3.  **环境检测**：外部集成（如 Git 同步）仅在检测到相应插件存在时才会启用。
+
+## 许可证
+MIT