原始内容
Byterover Skills
A collection of specialized Claude Code skills for enhanced knowledge management with Byterover MCP Server.
Overview
What is Byterover?
Byterover is an AI memory system that enables persistent knowledge storage and retrieval across conversations. It allows AI agents to:
- Store programming facts, patterns, and solutions
- Retrieve relevant context from past conversations
- Build cumulative knowledge over time
- Share knowledge across different projects
graph LR
A[AI Conversations] -->|Store| B[Byterover Memory]
B -->|Retrieve| C[Future Conversations]
C -->|Store| B
What are Byterover Skills?
Skills extend Byterover's core capabilities with specialized workflows:
- Structured documentation generation (reports, PRDs, architecture docs)
- Bidirectional integrations (Notion, GitHub, etc.)
- Advanced knowledge synthesis (aggregation, formatting, analysis)
- Automated workflows for common knowledge management tasks
graph TB
A[Byterover Core] --> B[byterover-notion-sync]
A --> C[Future Skills]
B --> D["Memories ↔ Notion"]
C --> E[Meetings, GitHub, Code]
Installation
Prerequisites
- Claude Code - Install from claude.ai/claude-code
- Byterover MCP Server - Set up from byterover.dev
- Notion MCP Server (optional) - Required for byterover-notion-sync (setup guide)
Installing Skills
Clone and copy skills to your Claude Code directory:
# Clone repository
git clone https://github.com/RyanNg1403/byterover-skills.git
cd byterover-skills
# Copy individual skill
cp -r byterover-notion-sync ~/.claude/skills/
# Or copy all skills
cp -r */ ~/.claude/skills/
Verify Installation
Start Claude Code and ask: "What Byterover skills are available?"
Skills
🔄 byterover-notion-sync
Bidirectional knowledge synchronization between Byterover memories and Notion documentation.
Core Capabilities:
- Transform memories → Notion documentation (4 formats: implementation reports, PRDs, feature docs, architecture docs)
- Import Notion pages → searchable memories with intelligent chunking
- Multi-query retrieval for comprehensive context
- Automatic format selection based on content
Architecture:
graph TB
subgraph mem2notion["Memories → Notion"]
A[Byterover Memories] -->|Multi-query| B[Aggregate Context]
B -->|Format Selection| C{Document Type}
C -->|Timeline-based| D[Implementation Report]
C -->|Requirements| E[PRD]
C -->|Usage Guide| F[Feature Docs]
C -->|System Design| G[Architecture Docs]
D --> H[Notion Page]
E --> H
F --> H
G --> H
end
subgraph notion2mem["Notion → Memories"]
I[Notion Page] -->|Fetch| J[Analyze Structure]
J -->|Chunk| K[Logical Sections]
K -->|Format| L[Memory Templates]
L -->|Store| M[Byterover Memories]
end
Example Use Cases:
"Create an implementation report on the auth system from July 10-15"
"Turn our notification discussions into a PRD"
"Document the microservices architecture"
"Import API documentation from Notion into memories"
Documentation:
- SKILL.md - Complete workflow guides
- references/ - Format templates and guidelines
Workflow Overview:
sequenceDiagram
participant U as User
participant C as Claude
participant B as Byterover
participant N as Notion
Note over U,N: Memories to Notion
U->>C: Create implementation report
C->>B: Multi-query memories
B-->>C: Aggregate results
C->>C: Select format and structure
C->>N: Create Notion page
N-->>U: Return page link
Note over U,N: Notion to Memories
U->>C: Import Notion page
C->>N: Fetch page content
N-->>C: Return structured content
C->>C: Chunk and format
C->>B: Store memories
B-->>U: Confirm storage
Usage
Skills activate automatically based on your requests - no special commands needed.
Example Workflows:
# Generate documentation from memories
User: "Create an implementation report on the payment system we built last week"
→ Queries memories → Structures as report → Creates Notion page with timeline
# Import documentation into memories
User: "Import our API documentation from Notion into memories"
→ Fetches Notion page → Breaks into chunks → Stores as searchable memories
# Architecture documentation
User: "Document our microservices architecture based on recent work"
→ Retrieves architecture memories → Structures with diagrams → Creates Notion doc
Contributing
Adding New Skills
- Fork this repository
- Create skill following this structure:
your-skill-name/ ├── SKILL.md # Required: Workflow instructions ├── references/ # Optional: Templates and guides ├── scripts/ # Optional: Executable helpers └── assets/ # Optional: Resources - Update this README with skill description
- Submit pull request
See CONTRIBUTING.md for detailed guidelines.
Roadmap
Planned Skills:
- byterover-github-integration - Sync PR discussions, code comments, and issues with memories
- byterover-meeting-notes - Convert transcripts into structured memories and action items
- byterover-code-knowledge - Extract code patterns, architectural decisions, and best practices
- byterover-research-synthesis - Aggregate memories into research reports and literature reviews
Have a skill idea? Open an issue or contribute!
License
MIT License - see LICENSE for details.
Acknowledgments
- Built for Claude Code by Anthropic
- Powered by Byterover MCP Server
- Integrations with Notion MCP