retro

Generalize recurring patterns from accumulated logs to refine agent definitions.

8|Updated Nov 4, 2016
One-click install
npx skills add https://github.com/coil398/dotfiles --skill retro-coil398
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: retro
Source: https://github.com/coil398/dotfiles/tree/main/.claude/skills/retro
Command: npx skills add https://github.com/coil398/dotfiles --skill retro-coil398

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Generalizes patterns from accumulated logs to improve agent definitions by orchestrating retrospector-driven workflows for structured reflection and continuous improvement.

Core Features & Use Cases

  • Orchestrates retrospector as the main Claude workflow to perform pattern generalization and agent-definition refinement.
  • Supports a meta-self-improvement mode to update workflow skeletons and generate actionable recommendations for back-end changes.
  • Use cases include reviewing past agent performance, extracting recurring patterns, and evolving agent definitions for better automation.

Quick Start

Run retrospector on your target project to begin the retrospective workflow.

Frequently Asked Questions about retro

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

FAQPage Schema
How do I generalize recurring patterns from accumulated logs to refine agent definitions?

It uses a retrospector sub-agent invoked via the Agent tool to analyze memory paths and extract recurring patterns, driving systematic improvements in agent behavior and task execution.

Can I automate retrospective reviews across multiple projects using memory paths?

Yes, retrospective reviews across multiple projects are supported by resolving proper memory paths and applying a flag-controlled meta-improvement workflow to ensure safe operation.

Do I need a specific sub-agent to perform pattern generalization and workflow improvement?

Yes, pattern generalization requires the retrospector sub-agent invoked via the Agent tool to safely orchestrate structured reflection and update workflow skeletons.

What is meta-self-improvement mode for updating workflow skeletons?

Meta-self-improvement mode is a flag-controlled workflow that updates workflow skeletons and generates actionable recommendations for back-end changes to evolve agent definitions.

Are there limitations when applying retroactive reviews to agent definitions?

The main limitation is that the meta-improvement workflow is flag-controlled and requires proper memory path resolution, meaning it cannot operate safely without the retrospector sub-agent.