evolution

name	evolution
slug	evolution
version	1.1.0
homepage	https://github.com/EricArcha/EVO-skill
changelog	Add defer option, execution constraints, installed skill optimization, and OpenClaw compliance gates
description	EN: Distill reusable patterns from workspace history into new skills with human-in-the-loop approval. ZH: 从工作区历史中提炼可复用模式，生成新的 Skills，全程由你掌控。
metadata	{"clawdbot":{"emoji":"🧬","requires":{"bins":[]},"os":["linux","darwin","win32"]}}

Evolution Skill

Scans all OpenClaw workspaces, distills reusable patterns, and generates skill candidates for your review.

When to Use

Run when you want to:

• Identify reusable patterns from recent workspace activity
• Generate new skills based on demonstrated workflows
• Find cross-workspace issues (docs in wrong location, knowledge silos)
• Improve automation based on past successes

Usage

Command	Description
/evolution	Run evolution scan (incremental mode, last 30 days)
/evolution --full	Full rescan (ignores last run, rescans everything)
/evolution --config	Edit evolution configuration
/evolution --setup-cron	Interactive cron setup
/evolution --help	Show help

First Run

On first run, you will be prompted:

1. "是否扫描过去90天建立基线？" — Scan last 90 days to establish baseline?
2. "是否设置定时运行cron？" — Set up automatic runs?

Workflow

Phase 1: Scan
- Scan configured workspaces (memory/, knowledge/, skills/)

Phase 2: Pattern Recognition
- Identify patterns across 4 categories:
  - 工作边界类 (Cross-workspace boundary issues)
  - 经验复用类 (Reusable prompt templates, workflows)
  - 效率提升类 (Repetitive operations that can be automated)
  - 系统健康类 (Memory bloat, stale knowledge)

Phase 3: Quality Filtering
- Apply frequency, reward, stability, boundary, and OpenClaw compliance gates

Phase 4: Generate Report
- Generate distillation report with candidate skills

Phase 5: User Review
- You review and confirm each candidate (Confirm/Defer/Reject/Modify)

Phase 6: Generate Staging
- On confirmation, skill generates staging SKILL.md

Phase 7: Manual Install
- You install manually (strict confirmation each time)

Phase 8: Track Installed Skills
- Track evolution-generated skills for optimization opportunities

Output Location

• Reports: ~/.openclaw/skills/evolution/reports/<date>-report.md
• Staging candidates: ~/.openclaw/skills/evolution/staging/<candidate-name>/

Configuration

Edit ~/.openclaw/skills/evolution/config.yaml:

# Workspaces to exclude from scanning
exclude_workspaces:
  - workspace-incognito

# Default scan time span (days)
scan_span_days: 30

# Baseline retention period (days)
baseline_retention_days: 90

# Cron schedule (leave empty to disable)
cron_schedule: ""

Integration with Weekly Summary

This skill integrates with weekly summary (周汇报) if you have it configured.

If you have weekly summary:

• Evolution status appears automatically in your weekly report
• You'll be prompted if evolution hasn't run this week

If you don't have weekly summary:

• Run /evolution manually whenever you want
• Or set up a cron: /evolution --setup-cron
• Evolution works standalone without any other skills

Strictly No Auto-Execution

This skill NEVER:

• Auto-creates skills without confirmation
• Auto-installs skills
• Modifies system configuration
• Deletes or moves files
• Overwrites existing skills

Every action requires your explicit confirmation.

Quality Gates

Each skill candidate must pass ALL gates:

Basic Gates (must pass)

Gate	Requirement
Frequency	Pattern appears ≥3 times in scan period
Reward	Measurable time savings or error reduction
Stability	Not a one-time/adhoc pattern
Boundary	Doesn't overlap with existing skills

Execution Constraint Gate (must pass)

Gate	Requirement
Verifiable	Must include a concrete verification method — how do we test this skill actually works?
Trigger Clarity	Must specify exact trigger conditions, not vague "remind me to..."
Actionable	Must define concrete actions, not just descriptions
Feedback Loop	Must define how the skill confirms it executed correctly

OpenClaw Compliance Gate (must pass)

Gate	Requirement
Architecture Fit	Does this follow OpenClaw best practices? (e.g., knowledge centralized, rules in proper locations)
Tool Alignment	Can this actually be executed with OpenClaw's tool set?
Context Appropriate	Is the scope appropriate for a skill, not too broad or too narrow?

If a candidate fails Execution Constraint Gate or OpenClaw Compliance Gate:

• Mark as "需要重构" (needs refactoring)
• Provide specific feedback on what must change
• Do not generate staging SKILL.md until refactored

User Response Options

For each candidate, reply:

• 确认 (confirm) → Generate formal SKILL.md in staging/
• 待定 (defer) → Include in next evolution run, don't forget
• 拒绝 (reject) → Record rejection reason (for future filtering)
• 修改 (modify) → Tell me what needs to change

Phase 8: Installed Skill Tracking

For skills that evolution generated and you installed:

1. Track execution: Note when the skill is triggered and its effectiveness
2. Flag issues: If the skill fails to constrain behavior, generate an optimization candidate
3. Report: Include installed skill performance in next evolution report

Optimization trigger conditions:

• Skill rarely triggered → low value, consider removing
• Skill triggered but ineffective → needs refactoring
• OpenClaw architecture changed → skill may be outdated

Fallback for Non-周汇报 Users

If you don't have a weekly summary skill:

1. Manual mode: Run /evolution whenever you want
2. Cron mode: Set up /evolution --setup-cron for bi-weekly runs
3. Flexible: Works completely standalone