{"skill":{"slug":"openclaw-agent-cost-guard","displayName":"OpenClaw Cost Guard","summary":"This skill should be used when the user asks to reduce OpenClaw token spend, audit model and cron cost risk, prevent denial-of-wallet incidents, add budget g...","description":"---\nname: openclaw-cost-guard\ndescription: This skill should be used when the user asks to reduce OpenClaw token spend, audit model and cron cost risk, prevent denial-of-wallet incidents, add budget guardrails, review expensive agent defaults, or tighten AI-agent cost governance before wider rollout.\n---\n\n# OpenClaw Cost Guard\n\nUse this skill to identify obvious cost leaks before they turn into a denial-of-wallet problem.\n\n## Goal\nMove from “the agents work” to “the agents work within an intentional budget.”\n\n## Workflow\n1. Identify the target OpenClaw config or workspace.\n2. Run the bundled cost-guard script.\n3. Review expensive defaults, missing budgets, and high-risk patterns.\n4. If the result is `FAIL`, stop scaling usage until cost controls are tightened.\n5. Apply the narrowest guardrails that reduce spend without breaking the workflow.\n\n## Command\n```bash\nnode {baseDir}/scripts/cost-guard.mjs --config ~/.openclaw/openclaw.json\n```\n\nThe `--config` flag is optional. If omitted, the script checks the default OpenClaw config path.\n\n## What the script checks\n- whether explicit budget signals exist\n- whether default models look expensive for always-on usage\n- whether browser/interactive tooling appears enabled without cost discipline\n- whether token limits appear excessively large\n- whether the config contains multiple high-cost patterns at once\n- whether the setup needs a governance recommendation before scaling\n\n## Output format\nThe script returns JSON with:\n- `score`\n- `verdict`\n- `summary`\n- `findings`\n- `recommendations`\n- `guardrails`\n- `evidence`\n\n## Verdicts\n- `PASS` — no major lightweight cost-governance gap found\n- `WARN` — spend risk exists and should be reviewed\n- `FAIL` — denial-of-wallet risk is materially elevated\n\n## Important limits\n- This is a lightweight static review, not a bill-reconciliation system.\n- A low-risk config can still become expensive through user behavior or external automations.\n- Always verify against real provider invoices and usage telemetry.\n\n## References\n- `{baseDir}/references/cost-playbook.md`\n","tags":{"latest":"0.1.0"},"stats":{"comments":0,"downloads":544,"installsAllTime":21,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1773546279293,"updatedAt":1778491917527},"latestVersion":{"version":"0.1.0","createdAt":1773546279293,"changelog":"Initial release: lightweight OpenClaw config audit for budget gaps, premium-model defaults, oversized token ceilings, and denial-of-wallet risk signals.","license":"MIT-0"},"metadata":null,"owner":{"handle":"x-rayluan","userId":"s17c0n0ce78mre7mrytgzehcwn83hkp1","displayName":"X-RayLuan","image":"https://avatars.githubusercontent.com/u/150766707?v=4"},"moderation":null}