| name | trae-claw-install |
| description | Run a repository-driven OpenClaw deployment workflow with platform routing, setup/start/check, acceptance checks, and unified troubleshooting. |
TraeClawInstall
Description
This skill executes an end-to-end OpenClaw deployment workflow from a repository that already provides platform scripts and operational documents.
Usage Scenario
Use this skill when:
- The user asks for one-stop OpenClaw local deployment.
- The user wants to follow a repository-standard install/start/check flow.
- The user has setup/start failures and needs structured troubleshooting.
Preconditions
- Current working directory is the repository root.
- Terminal execution is available.
- Repository contains platform scripts and troubleshooting docs.
Instructions
- Detect platform and route scripts:
- Windows: prefer WSL2 and use
scripts/windows/wsl/*.sh
- macOS: use
scripts/macos/*.sh
- Linux: use
scripts/linux/*.sh
- Validate baseline:
node --version major version is >=22
npm --version is available
- if
openclaw --version is missing, continue with setup step
- Execute standard flow:
- Run minimum acceptance:
openclaw doctor
openclaw status
openclaw dashboard
- If any step fails, run troubleshooting workflow:
- capture the first error line
- rerun doctor/status
- verify binary paths and versions
- apply fixes from repository troubleshooting docs
- change one variable at a time and re-verify
Output Contract
- Success: platform, executed steps, acceptance results, and service accessibility.
- Failure: first error, diagnostics performed, and next actionable fix.
Constraints
- Reuse repository scripts and docs; do not create a parallel flow.
- Never write real secrets; only use example configuration.
- On Windows, prefer execution inside WSL2 Linux filesystem.