self-improvement

To log errors and feature requests to Markdown for cross-session memory, you can use a self-improvement Skill that auto-detects and writes entries to dedicated `.learnings/ERRORS.md` and `FEATURE_REQUESTS.md` files.

Capturing AI agent learnings and insights involves logging corrections, knowledge gaps, and best practices into a structured `LEARNINGS.md` file within a `.learnings` directory to establish a living record for continuous improvement.

You need a `.learnings` directory at your project root containing `LEARNINGS.md`, `ERRORS.md`, and `FEATURE_REQUESTS.md` templates, which the logging system initializes and manages to guide the capture of learnings and errors.

Yes, the continuous improvement logging system supports openclaw integration, allowing captured learnings, errors, and feature requests stored in Markdown to be reused and promoted across different AI agents and projects.

Using a dedicated Markdown logging system provides structured templates for learnings, errors, and feature requests, enabling cross-session memory, skill promotion, and organized knowledge retrieval that standard text files lack.