From superpowers-memory
Use when initializing project knowledge for the first time or when knowledge has drifted too far from reality — full codebase scan and knowledge regeneration
How this skill is triggered — by the user, by Claude, or both
Slash command
/superpowers-memory:rebuildThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Scan the entire codebase and generate a complete project knowledge base from scratch.
Scan the entire codebase and generate a complete project knowledge base from scratch.
Announce at start: "I'm rebuilding the project knowledge base from the codebase."
Scan the codebase:
ls, key directories, entry pointspackage.json, Cargo.toml, pyproject.toml, Makefile, docker-compose.yml, etc.README.md, CLAUDE.md, docs/ directorydocs/superpowers/specs/, docs/superpowers/plans/git log --oneline -20Generate knowledge files:
Create docs/project-knowledge/ directory if it doesn't exist.
For each of the 5 knowledge files, use the plugin template as the structural basis and fill in concrete content from the codebase analysis:
Set frontmatter: For every generated file:
last_updated: today's date (YYYY-MM-DD)updated_by: superpowers-memory:rebuildtriggered_by_plan: "none" (literal string — not YAML null)Generate MEMORY.md index:
After writing the 5 knowledge files, generate docs/project-knowledge/MEMORY.md:
templates/MEMORY.md, setting updated_by: superpowers-memory:rebuild and triggered_by_plan: "none"Size constraint: Keep MEMORY.md under 30 lines total.
Commit:
git add docs/project-knowledge/
git commit -m "docs: rebuild project knowledge base from codebase"
superpowers-memory:update after the next iteration to keep knowledge freshnpx claudepluginhub jacexh/skill-workshop --plugin superpowers-memoryLoads project-specific context from a knowledge base document to inform all skills about the project's tech stack, architecture, and conventions.
Analyzes source code to discover and extract implicit knowledge—architecture patterns, conventions, API contracts, config structures, codebase rules—into KB articles.
Orchestrates parallel knowledge base generation from codebases using spatial analysis, map-reduce architecture, and sub-agents for concepts, architecture, modules, and patterns. Supports feature learning from archived features.