DMO Geek
← Back to all skills

Runtime Optimization

Smart Agent Optimizer

Tune OpenClaw runtime behavior for cost, speed, and context hygiene

Built for OpenClaw operators and power users

Smart Agent Optimizer applies safe, repeatable OpenClaw config patches to reduce token usage, limit stale context drift, and keep threaded sessions responsive over time.

$5

What It Does

  • • Applies context pruning in cache-ttl mode to limit stale context carryover.
  • • Configures thread idle reset windows to keep long-running threads clean.
  • • Provides practical presets (Aggressive, Balanced, Long Context) for common operating profiles.
  • • Uses patch-only config workflow with explicit completion notes and verification behavior.
  • • Supports targeted partial patches when you want to change only one setting.

How It Works

  • • Reads active config schema and runtime config before making any change.
  • • Selects preset values for contextPruning ttl and thread idle reset behavior.
  • • Applies a scoped `config.patch` payload with a human-readable note.
  • • Confirms the restart/config state and reports effective values back clearly.
  • • Never uses destructive apply flows for this workflow.

Setup Overview

  • • Purchase Smart Agent Optimizer from this page to unlock instant download.
  • • Extract/copy into ~/.openclaw/workspace/skills/smart-agent-optimizer/
  • • Ensure SKILL.md is present at ~/.openclaw/workspace/skills/smart-agent-optimizer/SKILL.md
  • • Start OpenClaw and ask: "Use Smart Agent Optimizer with the Balanced preset."
  • • Confirm your agent reports applied ttl and idle reset values after patching.

Key Files Included

SKILL.md

Optimization workflow, presets, patch template, and guardrails.

smart-agent-optimizer.skill

Packaged skill artifact for direct installation.

Real Use Cases

Challenge

Thread sessions keep carrying stale context and token usage is climbing.

Tool Application

Apply the Balanced preset (cache-ttl 15m, idle reset 120m) with Smart Agent Optimizer.

Outcome

Lower average context load and more consistent response quality in ongoing threads.

Challenge

A team wants stricter cost controls during heavy experimentation.

Tool Application

Use the Aggressive Cost Save preset to shorten ttl and idle windows.

Outcome

Reduced spend with controlled tradeoffs on long-context continuity.

Challenge

An ops lead only wants thread idle reset adjusted, nothing else.

Tool Application

Run a scoped patch for thread reset settings only.

Outcome

Targeted change without touching unrelated runtime config.

What's Included

  • • smart-agent-optimizer/SKILL.md
  • • smart-agent-optimizer.skill (packaged artifact)

Format & Compatibility

  • • Format: Markdown skill + packaged .skill artifact
  • • Runtime: OpenClaw with gateway config actions available
  • • Best for: Cost tuning, context hygiene, and runtime responsiveness control

Built on OpenClaw

This is a new service built on OpenClaw, an open-source AI agent framework.

These skills are actively maintained and updated based on user feedback.

Technical users can review OpenClaw documentation at docs.openclaw.ai and source code on GitHub.

Early Access

Join the first DMO marketers using agent-powered workflows.

Get Smart Agent Optimizer

Price: $5

Purchase opens the equivalent ClawMart listing to complete checkout for this skill.

Need help applying the right preset? Contact support via DMO Geek.