From lisa
Extract semantic memory from project analysis. Scans codebase, docs, and configs to understand tech stack, constraints, and goals.
How this command is triggered — by the user, by Claude, or both
Slash command
/lisa:discoverThe summary Claude sees in its command listing — used to decide when to auto-load this command
# Discover (Stage 1) - Lisa Scan the current project and generate Gastown-compatible semantic memory. ## When to Use Use this command when: - Starting migration of an existing project - Need to understand a codebase's tech stack - Want to document project constraints and goals - Preparing for roadmap generation ## What This Does 1. **Scans package files** - Detects runtime, framework, dependencies 2. **Reads configuration** - Identifies database, auth, deployment targets 3. **Analyzes documentation** - Extracts constraints, personas, non-goals 4. **Generates memory** - Outputs `.gt/mem...
Scan the current project and generate Gastown-compatible semantic memory.
Use this command when:
.gt/memory/semantic.jsonskills/discover/SKILL.md.gt/memory/ directory if it doesn't existsemantic.json with discovered factsevidence.files_analyzedCreates/updates .gt/memory/semantic.json with:
semantic.json is valid JSONproject.name is populatedtech_stack fields are populatedevidence.files_analyzed has at least 1 entry/lisa:discover
After running, check .gt/memory/semantic.json for results.
This command replaces the deprecated /lisa:analyze command. The functionality is identical, but the naming better reflects the staged architecture.
npx claudepluginhub auge2u/lisa-helps-ralph-loops --plugin lisa2plugins reuse this command
First indexed Jan 28, 2026
/initScans the current project to detect its technology stack (framework, language, database, ORM, auth, styling) and saves the findings as Nemp memories for future context.
/discoveryExplores the project codebase and documentation, identifies gaps, and asks targeted questions to clarify requirements before implementation.
/initAnalyzes the project and ingests all details into CORE memory with a hierarchical structure covering project foundation, technical context, user preferences, and active progress.
/initAnalyzes the project to build or update a shared context file for all agents. Detects tech stack, project structure, and key patterns, with optional --update flag for incremental refresh.
/output-templateGenerates complete YAML project fingerprint for Stage 0 analysis, including metadata, tech stack, architecture/AI/business hypotheses, and scanned files summary.
/onboardScans a new codebase, maps its architecture, sets up project rules, and saves context to memory for future sessions.