Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
Bootstraps a new OpenClaw instance with aligned model routing, consistent gateway tokens, anti-stall contracts, autonomy loop, and verification gates for sta...
Bootstraps a new OpenClaw instance with aligned model routing, consistent gateway tokens, anti-stall contracts, autonomy loop, and verification gates for sta...
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
Apply this skill to bootstrap a new OpenClaw instance into a stable, task-completing setup. Use this workflow in order: Align model routing and provider endpoint. Align gateway token values across runtime surfaces. Install anti-stall execution contract in AGENTS.md and HEARTBEAT.md. Enable lightweight periodic self-check. Run verification gates before calling setup complete. Keep changes minimal, auditable, and reversible.
Use one source of truth in ~/.openclaw/openclaw.json: { "models": { "mode": "merge", "providers": { "bailian": { "baseUrl": "https://coding.dashscope.aliyuncs.com/v1", "apiKey": "YOUR_CODING_PLAN_KEY" } } }, "agents": { "defaults": { "model": { "primary": "bailian/qwen3-coder-plus" } } } } Rules: Do not keep coding-intl.dashscope.aliyuncs.com in active config. Do not duplicate provider definitions in multiple files unless strictly needed. If ~/.openclaw/agents/main/agent/models.json exists, remove conflicting provider overrides.
Keep all token surfaces aligned: gateway.auth.token in openclaw.json gateway.remote.token in openclaw.json (if present) Dashboard Control UI token (paste the same value) PowerShell audit: $cfg = Get-Content "$HOME/.openclaw/openclaw.json" -Raw | ConvertFrom-Json $auth = $cfg.gateway.auth.token $remote = $cfg.gateway.remote.token "auth.token = $auth" "remote.token = $remote" if ($remote -and $auth -ne $remote) { Write-Warning "Token mismatch in openclaw.json" } After any token change: openclaw gateway restart
Write or update AGENTS.md with these mandatory constraints: Output state on each substantial task: Goal, Progress, Next. Do not stop before completion except for explicit blocker or user stop. On failure: retry, then fallback, then report minimal unblock input. Multi-step completion must include evidence artifact path or command result summary. Write or update HEARTBEAT.md: Each cycle performs at most 1-2 checks. Either produce execution evidence or exactly HEARTBEAT_OK. If queue item exists, execute one concrete step, then log evidence to memory/YYYY-MM-DD.md.
If cron/system events are available, create conservative jobs: Daily: unfinished-task check. Weekly: memory review and friction pattern extraction. If cron is unavailable, enforce manual equivalent: Start-of-day: review tasks/QUEUE.md, pick one actionable item. End-of-day: append one lesson to memory/YYYY-MM-DD.md.
Setup is complete only when all pass: Model call succeeds with primary route. Gateway access succeeds without token mismatch. A 3+ step task finishes with evidence. At least one blocker path is documented in memory/blocked-items.md. Recommended evidence: command summary artifact path or URL final status (done or blocked with required input)
Before edits, create backups: Copy-Item "$HOME/.openclaw/openclaw.json" "$HOME/.openclaw/openclaw.json.bak" -Force Rollback: Copy-Item "$HOME/.openclaw/openclaw.json.bak" "$HOME/.openclaw/openclaw.json" -Force openclaw gateway restart
For concrete command examples and troubleshooting playbooks, read: references/examples.md
Agent frameworks, memory systems, reasoning layers, and model-native orchestration.
Largest current source with strong distribution and engagement signals.