summarize
Compress without losing truth. Backlink to sources.
When & Why to Use This Skill
The Summarize skill is a specialized context optimization tool designed to compress large-scale information without losing essential truths. It enables AI agents to manage long-term memory by distilling session logs, research notes, and technical documents into concise summaries while maintaining strict traceability through mandatory backlinks to original sources. This ensures that the agent remains efficient and informed even when dealing with extensive datasets that exceed standard context window limits.
Use Cases
- Context Window Management: Automatically compress long-running session logs or research notebooks when the context budget is reached, preserving key insights while freeing up space.
- Knowledge Traceability: Generate structured summaries of complex project files that include direct backlinks, ensuring the AI can always reference the original source for implementation details.
- Information Tiering: Implement 'hot' and 'cold' hints within summaries to help the agent prioritize critical information and archive less relevant data without deleting it.
- Long-term Memory Archiving: Distill findings from a 'Research Notebook' into a 'Memory Palace' format, creating a permanent, searchable record of wisdom gained across multiple sessions.
| name | summarize |
|---|---|
| description | "Compress without losing truth. Backlink to sources." |
| license | MIT |
| tier | 1 |
| related | [honest-forget, session-log, memory-palace, play-learn-lift, research-notebook] |
| tags | [moollm, compression, knowledge, context, backlinks] |
Summarize
Compress without losing truth. Backlink to sources.
Context compression for memory management.
[!IMPORTANT] Always backlink. Every summary points to its source. Never orphan knowledge.
The Goal
When files are too large for context:
- Summarize — Extract key information
- Backlink — Point to original source
- Prioritize — Hot/cold hints for what matters
- Preserve — Never delete, just compress
Contents
| File | Purpose |
|---|---|
| SKILL.md | Full protocol documentation |
| SUMMARIES.yml.tmpl | Summary template |
Example
summary:
source: "designs/original-design.md"
created: "2025-12-30"
key_points:
- "Files are state, no hidden memory"
- "YAML comments carry meaning"
backlink: "../designs/original-design.md"
full_context_needed_for:
- "Implementation details"
- "Edge cases"
The Intertwingularity
Summarize enables the LIFT stage — compress wisdom for sharing.
graph LR
SUM[📝 summarize] -->|compresses| SL[📜 session-log]
SUM -->|compresses| RN[📓 research-notebook]
SUM -->|enables| HF[🌫️ honest-forget]
SUM -->|part of| PLL[🎮📚🚀 play-learn-lift LIFT]
SR[🔧 self-repair] -->|triggers| SUM
Dovetails With
Sister Skills
| Skill | Relationship |
|---|---|
| play-learn-lift/ | Summarize IS LIFT — share wisdom |
| honest-forget/ | Summarize before forgetting |
| session-log/ | Source material to compress |
| research-notebook/ | Findings to distill |
| self-repair/ | Triggers when context exceeds budget |
| memory-palace/ | Place summaries in palace rooms |
Protocol Symbols
| Symbol | Link |
|---|---|
SUMMARIZE |
PROTOCOLS.yml |
HONEST-FORGET |
PROTOCOLS.yml |
HOT-COLD |
PROTOCOLS.yml |
Kernel
- kernel/memory-management-protocol.md — Full specification
Navigation
| Direction | Destination |
|---|---|
| ⬆️ Up | skills/ |
| ⬆️⬆️ Root | Project Root |
| 🌫️ Sister | honest-forget/ |
| 🎮 Sister | play-learn-lift/ |