功能描述

Automates project inspection and iteration by analyzing from user, product, project, and tech perspectives to continuously improve code quality and delivery.

使用说明 (SKILL.md)

Auto-Evolve v4.4 (build 57fe0d7)

Name: Workspace
Author: relunctance

Four-perspective automated inspection and iteration manager.

Make your projects continuously better — automatically.

Core Philosophy

auto-evolve is not just a code scanner — it's a巡检伙伴 that thinks like a human.

On each scan, auto-evolve simulates receiving a Feishu message:

"What else can this project improve? Any shortcomings?"

It then examines the project from four perspectives, forming real opinions — not mechanically listing issues.

Scan Workflow (v4.0)

auto-evolve scan
    │
    ▼
┌─────────────────────────────────────────────────────┐
│  Step 1: project-standard project type detection      │
│  Detects: Skill / CLI / Python Library / Web / ...  │
│  Determines perspective weights + inspection focus     │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  Step 2: Four-perspective inspection               │
│                                                      │
│  👤 USER    → user/user-perspective.md (criteria) │
│  📦 PRODUCT → product-requirements.md (criteria)  │
│  🏗 PROJECT → project-inspection.md (criteria)     │
│  ⚙️ TECH   → code-standards.md (criteria)       │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  Step 3: project-standard reference docs            │
│  Used as evaluation criteria, output grouped report  │
└─────────────────────────────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│  Step 4: Execute / Notify / Record to learnings    │
└─────────────────────────────────────────────────────┘

Relationship with project-standard

Component	Role
project-standard	Defines taxonomy + four-perspective framework + reference docs (judging criteria)
auto-evolve	Loads standards, runs inspection, records learnings, executes improvements

Four-Perspective Framework

┌─────────────────────────────────────────────────────┐
│              auto-evolve Inspection Framework v4.0    │
├──────────────┬──────────────────┬───────────────────┤
│   User      │     Product      │     Project       │    Tech        │
│  "Usable?"  │ "Delivered?"    │   "Healthy?"     │  "Clean?"      │
├──────────────┼──────────────────┼───────────────────┼──────────────────┤
│ CLI design  │ Feature complete │ Learnings closed  │ Code quality   │
│ Learning    │ Promise kept     │ Scan history     │ Architecture  │
│ Errors      │ Pain resolved   │ Config rational  │ Test coverage  │
│ Fault tol.  │ Docs match code │ Dependency health│ Performance   │
└──────────────┴──────────────────┴───────────────────┴──────────────────┘

Four Perspectives Detail

👤 User Perspective

Core question: Is it pleasant to use?

Ask	Finds
CLI design	Non-intuitive flags, missing defaults
Learning curve	How long for a newcomer?
Error messages	Machine-speak vs human-speak
Fault tolerance	What on partial failure?
Workflow	Steps per operation?

📦 Product Perspective

Core question: Does it deliver what it promises?

Ask	Finds
README promises	Features claimed but not built
Pain points	❌-marked issues still broken
Feature completeness	Half-baked features
Docs consistency	Docs ≠ code

🏗 Project Perspective

Core question: Is it managed well?

Ask	Finds
Learnings loop	Previous findings tracked?
Scan rhythm	Regular schedule?
Config rationality	Over/under-configured?
Dependency health	Outdated deps? Known CVEs?

⚙️ Tech Perspective

Core question: Is the code healthy?

Ask	Finds
Code quality	Duplicates, long functions
Architecture	Module coupling
Test coverage	Core logic tested?
Performance/security	Bottlenecks, vulnerabilities

Note: Tech is the lowest priority — it's important but should not overshadow product truth.

Scan Output Format

🔍 auto-evolve Inspection Report — soul-force
Generated: 2026-04-05 22:30

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
👤 User Perspective ★★★★★
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  1. 🚨 Impact 0.7
     review command lacks --dry-run, users think it's safe but it writes files
     → Suggestion: Add --dry-run support to review

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📦 Product Perspective ★★★★
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  1. 🚨 Impact 0.8
     README promises "LLM fallback" but code has no fallback
     API failure = tool failure
     → Suggestion: Implement keyword-based rule engine as fallback

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚙️ Tech Perspective ★★
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
  [opt] 🟡 duplicate_code: SoulForgeConfig init repeated 15 times

Commands

scan

# Scan all configured repos
python3 auto-evolve.py scan

# Single repo scan
python3 auto-evolve.py scan --repo /path/to/repo

# Preview mode (no execution)
python3 auto-evolve.py scan --dry-run

# With specific persona memory
python3 auto-evolve.py scan --recall-persona master

confirm / reject / approve

python3 auto-evolve.py confirm
python3 auto-evolve.py reject 2 --reason "too risky"
python3 auto-evolve.py approve 1,3

repo-add / repo-list

python3 auto-evolve.py repo-add ~/.openclaw/workspace/skills/hawk-bridge --type skill
python3 auto-evolve.py repo-list

schedule

python3 auto-evolve.py schedule --every 168
python3 auto-evolve.py schedule --suggest

learnings

python3 auto-evolve.py learnings
python3 auto-evolve.py learnings --type rejections
python3 auto-evolve.py learnings --summary   # v4.3: summary view

trends (v4.3)

python3 auto-evolve.py trends --repo soul-force  # Scan trend for a project
python3 auto-evolve.py trends --all              # All projects

Configuration

~/.auto-evolverc.json

{
  "mode": "semi-auto",
  "full_auto_rules": {
    "execute_low_risk": true,
    "execute_medium_risk": false,
    "execute_high_risk": false
  },
  "schedule_interval_hours": 168,
  "repositories": [
    {
      "path": "/path/to/repo",
      "type": "skill",
      "visibility": "public",
      "auto_monitor": true
    }
  ]
}

LLM Integration

auto-evolve uses OpenClaw-configured LLM (no separate API key needed).

Priority: OPENAI_API_KEY / MINIMAX_API_KEY env vars, or openclaw config get llm.

Iteration Storage

.auto-evolve/
  .iterations/
    {id}/
      manifest.json        -- metadata + findings
      plan.md             -- execution plan
      pending-review.json -- items pending review
      report.md           -- execution report
      metrics.json        -- iteration metrics
  .learnings/
    approvals.json       -- approved changes
    rejections.json      -- rejected changes + reasons

安全使用建议

This package implements a powerful auto‑inspection + auto‑execution system that will read local OpenClaw/HawkBridge memory, run git/gh/pytest, commit & push changes, create releases, close issues, and can be scheduled to run automatically. Before installing, verify: (1) you trust the author (metadata is inconsistent), (2) there are no secrets or sensitive conversation data in ~/.openclaw or hawk-bridge you don't want read, (3) the 'gh' CLI and git credentials currently on the host grant only minimal scope (or run in a sandbox/test account), (4) run the code in a disposable/non-production repository first with mode set to semi-auto/dry-run, and (5) inspect and audit the bundled Python scripts (they are included) for any behavior you don't expect. If you are uncomfortable with automatic commits or issue-closing, do not enable full-auto or scheduling and consider rejecting this skill until metadata and required permissions are clarified.

功能分析

Type: OpenClaw Skill Name: hawk-bridge-v2 Version: 1.2.0 The 'auto-evolve' skill is a comprehensive project maintenance and automation tool for the OpenClaw ecosystem. It implements a 'Four-Perspective Framework' (User, Product, Project, Tech) to evaluate repositories using LLM-driven analysis. Key features include automated code refactoring (with syntax validation and git-based rollbacks), dependency analysis, test coverage comparison, and smart scheduling via OpenClaw's cron system. While the tool has extensive permissions—including file system write access, git management (commit/push/PR), and the ability to execute shell commands (git, pytest, gh)—these capabilities are strictly aligned with its documented purpose of continuous project improvement. The code is well-structured, includes safety checks, and lacks any indicators of malicious intent, data exfiltration, or obfuscation.

能力标签

crypto

能力评估

⚠ Purpose & Capability

The SKILL.md and code implement an Auto‑Evolve project-inspection/execution system (scanning repos, making/committing changes, creating releases, auto-closing GitHub issues, scheduling cron jobs, reading persona memories). That aligns with the description 'automates project inspection'. However registry metadata/inventory claims 'No required env vars' and 'instruction-only' while the package contains many Python scripts that call git/gh, access local OpenClaw SQLite and LanceDB, run pytest, and may call LLMs. The slug/owner metadata referencing 'hawk-bridge' is inconsistent with the skill name 'Workspace' and the code (auto-evolve), suggesting packaging/metadata mismatch.

⚠ Instruction Scope

SKILL.md and CLI examples instruct the tool to scan arbitrary repos, read persona memory (OpenClaw SQLite), read HawkBridge LanceDB, run tests, create/merge PRs, commit & push changes, create GitHub releases, and schedule recurring scans via 'openclaw cron'. These instructions require reading local files, credentials and mutating remote repos. The instructions also reference running commands that could transmit data (e.g., calling 'gh', pushing commits) and close GitHub issues automatically — all beyond a passive scanner.

ℹ Install Mechanism

No install spec is declared in the registry, but full Python source files are bundled in the package. There are no external downloads in the provided manifest (lower risk), but the code expects external CLIs/libraries (gh CLI, pytest, lancedb) and will call subprocesses. The mismatch (declared instruction-only vs included code) is an engineering inconsistency that deserves attention.

⚠ Credentials

The registry declares no required env vars or credentials, yet the code reads environment variables (OPENCLAW_AGENT_ID), accesses local OpenClaw workspace files and SQLite DBs, attempts to use HawkBridge LanceDB, and invokes 'gh' and 'git' via subprocess. Those tools rely on credentials or local state (GitHub token, SSH keys, stored tokens in gh). The code also refers to LLM configuration and cost tracking—all of which imply access to API keys/config not declared in metadata. Declaring no credentials is disproportionate to the actual access and operations.

⚠ Persistence & Privilege

always:false (good), but the tool is designed to modify repositories (auto-commit, push, create releases), close GitHub issues, create cron jobs via openclaw, and can be set to full-auto. Autonomous invocation is allowed by default; combined with the code's ability to modify remote repos, auto-close issues, and alter scheduling, this gives the skill broad potential impact. It also reads other skills' workspace/memory (OpenClaw/HawkBridge), crossing boundaries into other agents' data.

版本历史

v1.2.0

Add memory reliability system with forget/verify/correct UX

元数据

Slug hawk-bridge-v2

版本 1.2.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题

Workspace 是什么？

Automates project inspection and iteration by analyzing from user, product, project, and tech perspectives to continuously improve code quality and delivery. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 83 次。

如何安装 Workspace？

在 OpenClaw 或 Claude Code 对话框中运行命令「/install hawk-bridge-v2」即可一键安装，无需额外配置。

Workspace 是免费的吗？

是的，Workspace 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。

Workspace 支持哪些平台？

Workspace 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。

谁开发了 Workspace？

由 Gao.QiLin（@relunctance）开发并维护，当前版本 v1.2.0。

Workspace