retro

Conducts a retrospective on a coding session to suggest agent environment improvements.

4|Updated Jan 29, 2026
One-click install
npx skills add https://github.com/staticaland/skills --skill retro-staticaland
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: retro
Source: https://github.com/staticaland/skills/tree/main/plugins/retro/skills/retro
Command: npx skills add https://github.com/staticaland/skills --skill retro-staticaland

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? After a coding session with an AI agent, it is hard to know why the agent struggled, made mistakes, or wasted tokens. This Skill reviews session logs and turns those observations into concrete improvements for the agent's environment, such as navigation pointers, automated checks, and coding standards. ## Core Features & Use Cases - Session log analysis: Reads primary sources from a specified coding session (defaulting to the current one) to find friction points. - Categorized improvement candidates: Evaluates navigation, automated checks, coding standards, AGENTS.md bloat, tool economy, no-op instructions, and information access, then presents findings ordered by severity. - Environment file guidance: Applies clear rules for what belongs in CLAUDE.md/AGENTS.md versus CODING_STANDARDS.md versus docs, respecting the context pressure difference between implementation and review agents. - Use Case: After a long session where the agent repeatedly failed to find a config file and made a lint-catchable error, run a retrospective to get a navigation pointer added to AGENTS.md and a new lint rule proposed. ## Quick Start Ask the agent to run a retrospective on the current coding session and suggest improvements to the agent environment.

Frequently Asked Questions about retro

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

FAQPage Schema
How do I run a retrospective on an AI coding session?▼

Invoke the retro skill and optionally specify which session to review; it defaults to the current one. It reads the session logs, identifies improvement candidates across categories like navigation and automated checks, and presents them ordered by severity.

What kinds of improvements does a coding session retrospective suggest?▼

It suggests navigation pointers for hard-to-find files, automated checks like linting and tests for catchable errors, new or clarified coding standards for the reviewer agent, trimming of AGENTS.md, and fixes for token-inefficient tool calls.

Should coding standards go in AGENTS.md or a separate file?▼

Coding standards belong in CODING_STANDARDS.md, which is read during review, not implementation. AGENTS.md should stay minimal, usually only navigation pointers, because the implementation agent has the most context pressure.

When should I not use a session retrospective?▼

Skip it when the session had no notable friction, mistakes, or expensive tool calls, since there would be no improvement candidates. It is most valuable after long sessions, repeated agent errors, or when steering files have grown large and unwieldy.