learnings-researcher

You can surface past institutional learnings by searching docs/solutions using frontmatter metadata to match categories, modules, and symptoms to the current feature. The search ranks results by relevance and severity to prevent repeated mistakes and rework.

Searching by frontmatter metadata narrows results by matching specific fields like module, problem_type, component, and symptoms. This filtering ranks documented solutions by severity and returns concise summaries including root cause and recommended actions.

The best way to grep documented knowledge is to filter docs/solutions by matching frontmatter fields like module, problem_type, and symptoms. This approach ranks relevant solutions by severity and retrieves concise summaries of root causes and recommended actions.

Yes, you can retrieve recommended actions and root cause summaries. The search returns concise summaries from docs/solutions that include module, problem_type, symptoms, root_cause, and recommended actions matched by frontmatter metadata.

Matching frontmatter severity and categories reduces repeated development mistakes by surfacing relevant institutional learnings before coding begins. Ranking results by severity ensures the most critical problems and solutions are reviewed first.

Concise summaries of surfaced learnings include the module, problem_type, symptoms, root_cause, and recommended actions. These fields are extracted from the frontmatter metadata of matching documents in the docs/solutions directory.