gsd-extract-learnings

Automated learning extraction parses completed phase artifacts like PLAN.md and SUMMARY.md to capture decisions, lessons, and patterns. It generates a structured LEARNINGS.md file to streamline project review documentation.

Generating a LEARNINGS.md file requires completed phase artifacts such as PLAN.md, SUMMARY.md, VERIFICATION.md, UAT.md, and STATE.md. These documents provide the source content for extracting decisions and discovered patterns.

Documenting project learnings across multiple phases can be automated by running the extraction script against each phase's artifacts. This bypasses manual review and compiles structured knowledge documentation efficiently.

Knowledge documentation extraction requires no external dependencies, but it does require a workflow script for execution. It operates independently to parse phase artifacts and output structured learning files.

The best way to capture surprises and lessons learned during a phase review is using automated extraction to parse phase artifacts. It identifies and structures unexpected outcomes alongside decisions and patterns into a unified file.

Automated extraction for project learning documentation should not be used when phase artifacts like SUMMARY.md or VERIFICATION.md are incomplete or missing, as the script relies on these files to generate accurate structured learnings.