Can I review changes before they are applied to my files?

Yes, using the --list flag allows you to perform a dry-run to see which lessons are pending without actually modifying any documentation.

How does it decide where to place new information?

It analyzes the category of the lesson (e.g., bug, improvement, anti-pattern) and maps it to specific sections like CRITICAL RULES, Best Practices, or Workflow steps.

Does this skill work independently?

No, it requires the plan-marshall:manage-lessons skill to be present, as it relies on that skill's storage and retrieval system for lesson data.

What is the primary purpose of this skill?

It automates the process of taking 'lessons' (insights from previous coding sessions) and writing them into your plugin's markdown documentation files to improve future AI performance.

Plugin Lessons Learned Applier

Name: Plugin Lessons Learned Applier
Author: cuioss

bycuioss

•

Learning & Documentation

Automates the integration of accumulated development insights, bug fixes, and best practices into plugin documentation.

The Plugin Lessons Learned Applier is a specialized utility for the Plan Marshall orchestration layer that bridges the gap between discovering insights and memorializing them. By querying unapplied lessons captured during development, this skill intelligently maps bug reports, performance improvements, and architectural patterns to the specific sections of commands, agents, and skills where they are most relevant. It ensures that Claude Code's future behavior is informed by past experience, reducing repetitive errors and continuously refining the reliability of AI-driven workflows through automated documentation maintenance.

Key Features

01Automated documentation updates for commands, agents, and skill definitions

02Context-aware placement of insights into relevant markdown sections

03Intelligent lesson categorization (Bug, Improvement, Pattern, Anti-pattern)

04Dry-run capability to preview unapplied lessons before modification

05Bulk processing mode to synchronize all pending lessons across a project

061 GitHub stars

Use Cases

01Updating a build command with new error-handling logic discovered during a failed CI run

02Incorporating team-wide architectural patterns into specific AI agent system prompts

03Automating the 'memorialization' phase of a development sprint post-mortem

Key Features

01Automated documentation updates for commands, agents, and skill definitions

02Context-aware placement of insights into relevant markdown sections

03Intelligent lesson categorization (Bug, Improvement, Pattern, Anti-pattern)

04Dry-run capability to preview unapplied lessons before modification

05Bulk processing mode to synchronize all pending lessons across a project

061 GitHub stars

Use Cases

01Updating a build command with new error-handling logic discovered during a failed CI run

02Incorporating team-wide architectural patterns into specific AI agent system prompts

03Automating the 'memorialization' phase of a development sprint post-mortem