- name
- claw-code
- description
- AI Agent Harness Runtime — TypeScript port of Claude Code's harness system. Use when: (1) user asks about claw-code, agent harness, or Claude Code internals; (2) user wants to run claw-code commands or query tool/command registries; (3) routing prompts to commands or tools; (4) bootstrapping agent sessions; (5) parity auditing against original TypeScript archive. Triggers on phrases like "claw-code", "harness runtime", "Claude Code port", "agent harness", "run turn loop", "route prompt".
claw-code — AI Agent Harness Runtime
TypeScript port of Claude Code's harness system, compiled and ready to run.
Quick Start
cd C:\Users\dae\.openclaw\workspace\skills\claw-code\scripts
node dist/main.js <command> [options]Or build from source:
cd C:\Users\dae\.openclaw\workspace\skills\claw-code\scripts
npm install
npm run buildAvailable Commands
Info & Audit
node dist/main.js summary # Render workspace summary
node dist/main.js manifest # Print workspace manifest
node dist/main.js parity-audit # Compare against archive
node dist/main.js setup-report # Startup/prefetch report
node dist/main.js command-graph # Show command graph
node dist/main.js tool-pool # Show tool pool
node dist/main.js bootstrap-graph # Show bootstrap stagesQuery Registries
node dist/main.js subsystems --limit 16 # List Python modules
node dist/main.js commands --limit 20 # List mirrored commands
node dist/main.js commands --query <keyword> # Search commands
node dist/main.js tools --limit 20 # List mirrored tools
node dist/main.js tools --query <keyword> # Search tools
node dist/main.js tools --simple-mode # Core tools only
node dist/main.js tools --no-mcp # Exclude MCP tools
node dist/main.js show-command <name> # Show one command
node dist/main.js show-tool <name> # Show one toolRuntime Execution
node dist/main.js route "<prompt>" --limit 5 # Route prompt to commands/tools
node dist/main.js bootstrap "<prompt>" --limit 5 # Bootstrap a session
node dist/main.js turn-loop "<prompt>" --max-turns 3 # Run turn loop
node dist/main.js flush-transcript "<prompt>" # Persist session
node dist/main.js load-session <session_id> # Load saved sessionRemote Modes
node dist/main.js remote-mode <target>
node dist/main.js ssh-mode <target>
node dist/main.js teleport-mode <target>
node dist/main.js direct-connect-mode <target>
node dist/main.js deep-link-mode <target>Execution
node dist/main.js exec-command <name> "<prompt>" # Execute mirrored command
node dist/main.js exec-tool <name> <payload> # Execute mirrored toolProject Structure
claw-code/
├── SKILL.md # This file
└── scripts/
├── dist/ # Compiled JavaScript (ready to run)
│ └── main.js # CLI entry point
├── src/ # TypeScript source
│ └── reference_data/ # Snapshot data
├── package.json
└── tsconfig.jsonArchitecture
- Runtime Core (
runtime.ts) — PortRuntime class: route_prompt, bootstrap_session, run_turn_loop - Tool/Command Registries — loaded from
reference_data/*.jsonsnapshots - Session Management — session_store.ts, history.ts
- Query Engine — query_engine.ts with streaming support
- Graphs — bootstrap_graph.ts, command_graph.ts
- Remote Modes — remote_runtime.ts, direct_modes.ts