| name | openclaw-agent-token-optimizer |
| description | Optimize an OpenClaw agent setup (model routing, context management, delegation, rules, memory). Use when asked about optimizing agents, improving OpenClaw setup, or agent best practices. Use when this capability is needed. |
| metadata | {"author":"demerzels-lab"} |
OpenClaw Agent Optimization
Use this skill to tune an OpenClaw workspace for cost-aware routing, parallel-first delegation, and lean context.
Workflow (concise)
- Audit rules + memory: ensure rules are modular/short; memory is only restart-critical facts.
- Model routing: confirm tiered routing (lightweight / mid / deep) aligns with live config.
- Context discipline: apply progressive disclosure; move large static data to references/scripts.
- Delegation protocol: parallelize independent tasks; use sub-agents for long/isolated work.
- Heartbeat optimization: treat heartbeat as control-plane; move heavy checks to isolated cron/scripts; offer profiles A/B/C and require user choice if removing checks.
- Safeguards: add anti-loop + budget guardrails; prefer fallbacks over retries.
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/memory-patterns.md
references/continuous-learning.md
references/safeguards.md
Converted and distributed by TomeVault — claim your Tome and manage your conversions.