OrgClaw - OpenClaw Skill

Name: orgclaw
Version: 0.1.0
Author: Conpera Team
License: MIT

Description

Organization knowledge federation system for OpenClaw. Automatically extracts, scores, and stores reusable experiences from Agent tasks.

Features

• Auto-Extract: Automatically capture experience on task completion
• Quality Scoring: 4-dimension quality assessment (completeness, specificity, actionability, reusability)
• Smart Storage: Save high-quality experiences to personal knowledge base
• Pattern Linking: Connect with conpera-patterns knowledge base
• Semantic Search: Vector-based experience retrieval

Installation

Method 1: One-line Install (Recommended)

curl -fsSL https://raw.githubusercontent.com/conpera/orgclaw/main/install.py | python3

Method 2: OpenClaw CLI

openclaw skill install conpera/orgclaw

Method 3: Manual

git clone https://github.com/conpera/orgclaw.git ~/.openclaw/skills/orgclaw
cd ~/.openclaw/skills/orgclaw
python3 install.py

Configuration

Default configuration in ~/.openclaw/config.yaml:

skills:
  orgclaw:
    enabled: true
    auto_extract: true
    quality_threshold: 0.4
    min_lines_changed: 5
    personal_dir: ~/.orgclaw/personal
    enable_patterns: true

Usage

Automatic (Recommended)

Once installed, OrgClaw automatically extracts experience when OpenClaw tasks complete:

[OpenClaw] Task completed successfully!
[OrgClaw] ✅ Experience auto-saved (quality: 0.51)
  💾 Saved to: ~/.orgclaw/personal/exp-xxx.json

Manual Commands

# View statistics
orgclaw stats

# Search experiences
orgclaw search "memory leak"

# Search patterns
orgclaw patterns "api" --category coding

# Manual extract
orgclaw extract task-001 --description "Fixed bug in..."

How It Works

1. Task Completion: OpenClaw completes an Agent task
2. Hook Trigger: post_task.py is called automatically
3. Experience Extraction: Parse task description, detect patterns
4. Quality Scoring: Score on 4 dimensions (each 25% weight)
5. Storage Decision: Save if quality >= threshold (default 0.4)
6. Pattern Linking: Suggest related patterns from conpera-patterns

Architecture

Agent Task
    ↓
OpenClaw Hook (post_task.py)
    ↓
ExperienceExtractor
    ├── Pattern Detection (bug_fix/refactor/feature/optimization)
    ├── Solution Extraction
    └── Lesson Generation
    ↓
ExperienceScorer
    ├── Completeness (25%)
    ├── Specificity (25%)
    ├── Actionability (25%)
    └── Reusability (25%)
    ↓
Decision Gate (quality >= 0.4?)
    ├── Yes → ~/.orgclaw/personal/
    └── No → Discard + Suggestions

Data Flow

• Input: Task description, git diff, changed files
• Process: Extract → Score → Decide → Store
• Output: Structured experience JSON
• Storage: ~/.orgclaw/personal/*.json

File Locations

Type	Path
Skill Code	~/.openclaw/skills/orgclaw/
Configuration	~/.openclaw/config.yaml
Personal Data	~/.orgclaw/personal/
Hook	~/.openclaw/skills/orgclaw/hooks/post_task.py

Integration

With conpera-patterns

OrgClaw automatically links with conpera-patterns:

• Extracted experiences suggest relevant patterns
• High-quality experiences can be promoted to patterns
• Bidirectional knowledge flow

With OpenClaw

Required OpenClaw support:

• post_task hook system
• Task result metadata (description, commit, changes)

Development

# Clone
git clone https://github.com/conpera/orgclaw.git
cd orgclaw

# Install dev dependencies
pip install -e ".[dev]"

# Run tests
pytest tests/

# Type check
mypy orgclaw/

Troubleshooting

Issue: Hook not called

Solution: Check if OpenClaw supports post_task hook. Use manual extraction:

orgclaw extract <task-id> --description "..."

Issue: Import errors

Solution: Reinstall dependencies:

pip install -r ~/.openclaw/skills/orgclaw/requirements.txt --force-reinstall

Issue: Permission denied

Solution: Use user installation:

python3 install.py --user

Contributing

1. Fork the repository
2. Create a feature branch
3. Make changes
4. Submit PR

See CONTRIBUTING.md for details.

Roadmap

• Auto-extract from tasks
• Quality scoring
• Pattern linking
• Web UI for browsing
• Team/organization sharing
• LLM-enhanced extraction
• Feedback learning

License

MIT License - see LICENSE

Links

• GitHub: https://github.com/conpera/orgclaw
• Patterns: https://github.com/conpera/conpera-patterns
• Issues: https://github.com/conpera/orgclaw/issues