An automated efficiency optimizer and self-evolution manager for AI agents that periodically analyzes token usage, declutters redundant capabilities, and recommends performance improvements.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-efficiency-manager
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agent-efficiency-manager using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Agent Efficiency Manager?
The Agent Efficiency Manager is a specialized tool within the ecosystem of Openclaw Skills, designed to maximize the performance and minimize the operational costs of your AI agents. By continuously scanning your local configuration, this tool calculates precise token metrics, assesses performance scores, and uncovers high-impact opportunities for optimization. It eliminates redundant agent capabilities and ensures that every active skill aligns perfectly with your production objectives.
With built-in capabilities for self-evolution, the manager automatically retrieves the latest community packages, conducts compatibility checks, and recommends strategic upgrades. This maintains your developer agents at peak efficiency without manual profiling, directly lowering API expenses while boosting task execution speeds.
Agent Efficiency Manager Use Cases
- When users demand agent optimization or seek to aggressively lower LLM token consumption costs.
- Setting up scheduled automation to routinely analyze agent parameters and send push alerts.
- Discovering and evaluating new, highly compatible tools from the hub to enhance specific agent setups.
- Compiling comprehensive diagnostic reports and cost-saving recommendations for technical stakeholders.
How Agent Efficiency Manager Works
- Agent Configuration Scanning: The tool parses the local config setup to map the distribution of registered skills across all active agents.
- Token & Efficiency Analysis: It runs diagnostic scripts to calculate estimated context tokens, check relevance rankings, and identify redundant or overlapping capabilities.
- Skill Discovery: The manager communicates with remote repositories to aggregate candidate packages and assess their role relevance.
- Recommendation Generation: It evaluates current needs against candidate upgrades, producing a secure patch file suggestion.
- Evolution and Notifications: It streams the reports to team channels via webhooks and tracks optimization adoption rates over time.
Agent Efficiency Manager Setup
Install the Agent Efficiency Manager and its dependencies to activate self-optimizing capabilities.
# Run the efficiency analyzer to audit your current agent configuration
python3 scripts/analyze_agent_efficiency.py --config ~/.qclaw/openclaw.json --output metrics.json
# Retrieve available options from the central skill repository
python3 scripts/fetch_skillhub_skills.py --output available_skills.json
# Generate tailored upgrade recommendations based on your unique setups
python3 scripts/generate_recommendations.py --config ~/.qclaw/openclaw.json --available available_skills.json --output recommendations.json
To set up automated weekly reports, integrate the push notification module with your chosen communication platform:
# Push suggestions directly to your internal channels (e.g., WeChat, Enterprise Chat)
python3 scripts/push_notifications.py --recommendations recommendations.json --channel wecom --webhook YOUR_WEBHOOK_URL
Agent Efficiency Manager Data Schema & Taxonomy
The Agent Efficiency Manager structures its evaluation telemetry and reporting files across a standardized taxonomy:
| Filename / Variable | Format | Description |
|---|---|---|
metrics.json |
JSON | Key metrics file containing active skill_count, estimated_tokens, and calculated efficiency_score. |
available_skills.json |
JSON | Cached index of remote additions fetched from the repository. |
recommendations.json |
JSON | Curated upgrade paths, cost savings, and configuration patches. |
metrics_history.json |
JSON | Ongoing historical log of tracked indicators for trend reporting. |
trend_report.md |
Markdown | Rendered analysis showing token consumption trends, adoption rates, and performance evolution over time. |
Metrics Schema Reference
skill_count: Total number of active capabilities registered for an agent.estimated_tokens: Cumulative baseline overhead context size (assumes ~500-1000 tokens per unique skill definition).efficiency_score: Comprehensive scoring index (0 to 100) evaluating role-to-skill alignment.redundancy_count: Total volume of overlapping or completely unrelated functional modules flagged for removal.
Agent Efficiency Manager Advanced Features
- Proactive Pattern Matching: Leverages codified refactoring strategies (such as deduplication, domain isolation, and baseline simplification) to safely trim system prompts.
- Multi-Agent Optimization: Analyzes deep contextual boundaries across multi-agent systems, advising on which specialized tasks should be separated or combined.
- Cron-Triggered Automated Audits: Easily schedules system diagnostics on headless servers via simple automation configurations for periodic delivery.
- Interactive Patch Previews: Generates declarative patch configurations (
config.patch), allowing developers to review code changes before altering vital system structures. - Self-Improving Recommendation Engine: Tracks long-term adoption dynamics to adjust future suggestions and optimize suggestion accuracy automatically.
SKILL.md
Loading
Related Openclaw Skills
Anthropic Founder Handbook (2026 Edition)
An expert framework and AI agent skill implementing the 2026 Anthropic startup methodology to shift founder focus from execution to strategic judgment.
Meta Service Lead Gen
A specialized AI agent skill designed to audit, diagnose, and optimize high-ticket service lead generation funnels running on Meta, Facebook, and Instagram ads.
Computer Efficiency Skill
An intelligent, cross-platform system diagnostic and resource optimization agent skill with strict safety boundaries and automatic rollbacks.
Industrial Fund Investment Advisor
An advanced primary-market investment analysis AI agent designed to simulate top VC, PE, and government-guided fund decision-making using dynamic weight frameworks.
Complaint Handling Skill
An automated enterprise workflow skill modeled after standard corporate management procedures to systematically receive, categorize, address, and track customer complaints.
enterprise-credit-check
An authoritative enterprise background check tool that performs a five-dimensional credit verification using exclusively official government data sources.
Featured*
