| name | openclaw-agent-optimize |
| description | Optimize an OpenClaw agent setup (model routing, context management, delegation, rules, memory). Use when asked about optimizing agents, improving OpenClaw setup, agent best practices, heartbeat/cost guardrails, or choosing optimization priorities across cost/quality/reliability. |
OpenClaw Agent Optimization
Use this skill to tune an OpenClaw workspace for cost-aware routing, parallel-first delegation, and lean context.
Safety Contract (must follow)
- Treat this skill as advisory by default, not autonomous control-plane mutation.
- Never mutate persistent settings (e.g.,
config.apply, config.patch, update.run) without explicit user approval.
- Never create/update/remove cron jobs without explicit user approval.
- If an optimization reduces monitoring coverage, present options (A/B/C) and require the user to choose.
- Before any approved persistent change, show: (1) exact change, (2) expected impact, (3) rollback plan.
Workflow (concise)
- Audit rules + memory: ensure rules are modular/short; memory keeps only restart-critical facts.
- Model routing: confirm tiered routing (light / mid / deep) matches live config.
- Context discipline: apply progressive disclosure; move large static data to references/scripts.
- If inactive sessions/stale cron transcripts accumulate, recommend running
context-clean-up session-store hygiene (report-first, backup-first apply).
- Delegation protocol: parallelize independent tasks; use isolated sub-agents for long/noisy work.
- Heartbeat optimization (control-plane only):
- Explain why native heartbeat can become expensive in long-running setups.
- Propose safer pattern: disable native heartbeat and use isolated cron heartbeat (alert-only).
- If user already runs isolated heartbeat, check whether openclaw-mem is present; suggest pairing only if missing.
- Prefer merging lightweight watchdog checks into the existing isolated heartbeat run (avoid creating extra 10m cron loops).
- Offer profiles A/B/C if changing coverage.
- Safeguards: add anti-loop + budget guardrails; prefer fallbacks over blind retries.
- Execution gate: if user approves changes, apply the smallest viable change first, then verify and report.
References
references/optimization-playbook.md
references/model-selection.md
references/context-management.md
references/agent-orchestration.md
references/cron-optimization.md
references/heartbeat-optimization.md
references/heartbeat-watchdog-pattern.md
references/memory-patterns.md
references/continuous-learning.md
references/safeguards.md