memory-summarization git:20260507.44b06c0A

Immutable. This exact content is served forever at /api/v1/blob/12bb4773bf2b56e4.

---
name: memory-summarization
description: Conversation summarization for memory compression and context management
allowed-tools:
  - Read
  - Write
  - Edit
  - Bash
  - Glob
  - Grep
graph:
  domains: [domain:software-engineering]
  specializations: [specialization:ai-agents-conversational]
  skillAreas: [skill-area:context-management, skill-area:llm-summarization-qa]
  roles: [role:ml-engineer, role:backend-engineer]
  workflows: [workflow:feature-development, workflow:ml-model-lifecycle]

---

# Memory Summarization Skill

## Capabilities

- Implement conversation summarization strategies
- Configure rolling summary updates
- Design hierarchical summarization
- Implement token-aware summarization
- Create extractive and abstractive summaries
- Design summary quality evaluation

## Target Processes

- conversational-memory-system
- long-term-memory-management

## Implementation Details

### Summarization Strategies

1. **Rolling Summary**: Update summary with new messages
2. **Hierarchical**: Multi-level summarization
3. **Token-Budget**: Fit within token limits
4. **Extractive**: Key message selection
5. **Abstractive**: LLM-generated summaries

### Configuration Options

- LLM for summarization
- Summary token budget
- Update frequency
- Summary template
- Quality thresholds

### Best Practices

- Balance detail vs compression
- Preserve key information
- Monitor summary quality
- Test with long conversations
- Handle context window limits

### Dependencies

- langchain-core
- LLM provider