summarize-scratchpads

Compress multi-agent scratchpad entries into a structured summary of tasks and blockers.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/bert-berkers/UrbanRepML --skill summarize-scratchpads
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: summarize-scratchpads
Source: https://github.com/bert-berkers/UrbanRepML/tree/main/.claude/skills/summarize-scratchpads
Command: npx skills add https://github.com/bert-berkers/UrbanRepML --skill summarize-scratchpads

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps you quickly understand the current state of multiple AI agents by summarizing their recent activities and interactions, saving you from reading through lengthy individual logs.

Core Features & Use Cases

  • Cross-Agent Summarization: Consolidates information from multiple agent scratchpads into a concise overview.
  • State Orientation: Provides a snapshot of ongoing tasks, blockers, tensions, and unresolved issues across agents.
  • Use Case: Before a daily stand-up meeting, use this Skill to get a rapid understanding of what each agent is working on, any impediments they face, and any inter-agent conflicts that need resolution.

Quick Start

Summarize all scratchpad entries from today and yesterday.

Frequently Asked Questions about summarize-scratchpads

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I summarize multi-agent state across multiple scratchpads?

To summarize multi-agent state, compress scratchpad entries from today and yesterday into a structured overview. This process identifies in-progress tasks, blockers, inter-agent tensions, and unresolved questions across all active agents.

What's the best way to track debugging progress and blockers for multiple AI agents?

Tracking debugging progress and blockers for multiple agents requires condensing their scratchpad logs into a single state summary. This approach surfaces unresolved questions, inter-agent tensions, and notable observations without reading individual logs.

Can I identify stale agents or missing scratchpads in a multi-agent workflow?

Yes, you can identify stale agents and missing scratchpads in a multi-agent workflow. The summarization process evaluates recent activity to flag agents with missing data, bloated scratchpads, or stale status based on inactivity.

How does cross-agent summarization handle inter-agent tensions and unresolved questions?

Cross-agent summarization handles inter-agent tensions by extracting and grouping conflicting state information from individual scratchpads. It actively identifies unresolved questions and highlights tensions that require manual resolution.

Does multi-agent state management require external dependencies to condense scratchpad logs?

No external dependencies are required to condense scratchpad logs for multi-agent state management. The skill operates independently using internal scripts and references to parse and structure agent activity data.

When do I need to condense scratchpad entries for a daily stand-up meeting?

You need to condense scratchpad entries for a daily stand-up meeting when multiple agents are running asynchronously. Summarizing today's and yesterday's logs provides a rapid snapshot of ongoing tasks and impediments for review.