solo-retro

What This Skill Does

The solo-retro skill serves as the formal post-mortem engine for OpenClaw AI agent pipelines. It is designed to act as an objective auditor that reviews the efficiency and output of automated workflows. Rather than focusing on simple code syntax, solo-retro focuses on 'process quality'—identifying where the agent spent time, why certain steps failed, and how to improve the overall agent trajectory. By parsing pipeline logs, iteration logs, and project metadata, it provides a structured feedback loop, turning failed or inefficient tasks into actionable configuration patches.

Installation

To integrate this auditing capability into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/fortunto2/solo-retro

Ensure your project contains the standard .solo/pipelines/ directory structure so the skill can locate necessary logs and status files.

Use Cases

• Pipeline Post-Mortem: Immediately after a long-running task completes, run solo-retro to identify if the agent took 'scenic routes' or got stuck in inefficient loops.
• Optimization Audits: Run the skill periodically to analyze recurring failure patterns in your development process, such as consistent test failures or CI bottlenecks.
• Fallback Diagnostics: Use it on legacy projects lacking pipeline logs to gain insights via git history, file structure analysis, and code quality heuristics.
• Knowledge Transfer: Generate a summary of 'what went wrong' to update your CLAUDE.md documentation, ensuring the agent learns from past mistakes.

Example Prompts

1. "The pipeline finished, but it felt slow. Can you run a /retro and tell me where the waste was?"
2. "/retro - Analyze the last pipeline run and suggest patches for my build scripts."
3. "What went wrong during the last migration? Please evaluate the pipeline logs and identify recurring error patterns."

Tips & Limitations

• Self-Containment: This skill is strictly self-contained. It will not delegate to other agents, ensuring a focused and consistent analysis. Do not expect it to perform actual code changes; it is an analytical engine that provides recommendations.
• Artifact Requirements: The quality of the retro depends on the availability of logs. If you are starting a new project, ensure your environment maintains the .solo directory structure for the best results.
• Fallback Mode: If logs are missing, the skill switches to high-level repository analysis (Git history and code metrics). This is less granular but still excellent for assessing project health.