Review
Persistent memory for Claude Code - powered by Graphiti knowledge graphsFrom the repository description
npx -y skills add TonyCasey/lisa --skill reviewAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 9 stars9 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
- runs commandsInstructs the agent to run 3 commands, including `lisa review run [--force]` and 2 more.
SKILL.md
2.9 KB, 801 tokens by cl100k_base, as published. Nobody here has run it
Review Skill
Purpose
Automatically analyzes a codebase when Lisa is installed, creating a foundational memory of the project structure, technologies, and patterns. This review serves as context for all future Claude sessions.
Triggers
Use when the user says things like:
- "run review"
- "analyze this codebase"
- "scan the project"
- "what is this project about"
- "refresh codebase summary"
How to use
Automatic (during npm install)
The review runs automatically when Lisa is installed via npm install @tonycasey/lisa. It:
- Detects if the folder is a codebase
- Runs static analysis (language, framework, structure)
- Stores result as first memory
- Queues background AI enrichment
Manual commands
# Run review (or re-run with --force)
lisa review run [--force]
# Show current review
lisa review show
# Check status (done, enriched, etc.)
lisa review status
I/O contract
Static analysis output
{
"status": "ok",
"action": "run",
"result": {
"version": "1.0",
"project": { "name": "lisa", "path": "/dev/lisa", "groupId": "dev-lisa" },
"codebase": {
"language": "TypeScript",
"languages": ["TypeScript", "JavaScript"],
"framework": null,
"buildTools": ["npm", "tsc"]
},
"structure": {
"entryPoints": ["src/cli.ts", "src/index.ts"],
"mainModules": ["src/domain/", "src/infrastructure/"],
"testDirs": ["tests/"]
},
"dependencies": {
"count": 12,
"noteworthy": ["commander", "fs-extra", "@anthropic-ai/claude-code"]
},
"patterns": {
"architecture": "clean-architecture",
"testing": "node-test"
},
"metrics": {
"fileCount": 45,
"hasTests": true,
"hasDocumentation": true
}
},
"summary": "TypeScript CLI project with clean-architecture pattern..."
}
Show output
{
"status": "ok",
"action": "show",
"review": "TypeScript CLI project with clean-architecture pattern...",
"enriched": true,
"timestamp": "2026-01-11T15:30:00Z"
}
Status output
{
"status": "ok",
"action": "status",
"done": true,
"enriched": true,
"timestamp": "2026-01-11T15:30:00Z",
"groupId": "dev-lisa"
}
Memory storage
Init reviews are stored with these tags:
type:init-review- Identifies as init review memoryscope:codebase- Full codebase analysisai:enriched- Present if AI enrichment completed
Marker file
Location: .lisa/.init-review-done
Prevents re-running on subsequent installs. Delete to force re-run, or use --force flag.
Cross-model checklist
- Codex: Use explicit CLI commands; verify triggers match
- Claude: Keep output concise; shown in session-start context
- Gemini: Explicit commands; avoid model-specific tokens
Notes
- Static analysis targets <2 seconds execution
- AI enrichment runs in background (30-60 seconds)
- Logs written to
.lisa/.init-review.log - Never blocks npm install - errors are caught and logged
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.