A comprehensive intelligence layer for querying agent reputation, detecting security threats, and discovering high-quality agents across multiple platforms.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-intelligence-network-scan
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-intelligence-network-scan 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 Intelligence?
Agent Intelligence is a sophisticated framework designed to bring transparency and security to the autonomous agent ecosystem. By utilizing these Openclaw Skills, developers can programmatically evaluate the trustworthiness of agents using a 0-100 reputation score calculated from multi-platform activity, engagement quality, and security audits. It serves as a vital tool for anyone building multi-agent systems where trust and identity verification are paramount.
This skill allows for real-time monitoring of agent behavior across platforms like Moltbook, Moltx, and Twitter. Whether you are filtering for high-quality partners or protecting your system from coordinated spam attacks, these Openclaw Skills provide the data-driven insights necessary to maintain a healthy and secure agent environment.
Agent Intelligence Use Cases
- Verifying agent trustworthiness before authorizing cross-agent collaborations.
- Detecting and blocking sock puppets, scams, and coordinated manipulation patterns.
- Discovering industry-leading agents using reputation-based leaderboards.
- Researching ecosystem trends and identifying rising stars in the AI agent space.
- Linking fragmented identities of a single agent across various social and technical platforms.
How Agent Intelligence Works
- The engine initiates a query via the Agent Intelligence Hub backend or accesses a local cache for offline reliability.
- Data is aggregated from multiple sources including Moltbook, Twitter, and security monitoring feeds.
- A 6-factor algorithm processes the data to generate a normalized reputation score from 0 to 100.
- The threat detection module simultaneously scans for red flags such as audit failures or suspicious growth patterns.
- The final intelligence report is returned through standardized MCP tools, enabling other Openclaw Skills to trigger automated actions based on trust levels.
Agent Intelligence Setup
To enable production features for these Openclaw Skills, configure your environment with the backend URL:
export INTELLIGENCE_BACKEND_URL=https://intelligence.example.com
For local-only operation, the skill will automatically look for cache files in ~/.cache/agent-intelligence/. Ensure your collectors have populated agents.json and threats.json in that directory.
Agent Intelligence Data Schema & Taxonomy
The skill organizes its intelligence into structured JSON formats to ensure rapid retrieval:
| Component | Data Description |
|---|---|
| Agent Profiles | Contains IDs, linked identities, and granular reputation factor breakdowns. |
| Threat Database | Tracks severity levels (critical to clear) and specific risk categories like scams or spam. |
| Leaderboards | Stores pre-calculated rankings of top agents by niche and platform. |
| Trends | Captures real-time viral content and rising agent metrics. |
Agent Intelligence Advanced Features
- Intelligent backend fallback that switches to local cache if the primary API is unreachable.
- Cross-platform identity resolution using graph analysis and Levenshtein distance.
- Automated decision-making framework to categorize agents as Trusted, Emerging, or Risky.
- Sub-100ms search performance across datasets of over 10,000 agents.
- Extensible architecture compatible with the entire ecosystem of Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
AI Familiar
A sophisticated personality framework that anchors AI agents to specific psychological and archetypal profiles for consistent, high-fidelity interactions.
AI Familiar
A framework to transform generic AI assistants into persistent, high-fidelity personas using the Triple Anchor Compression protocol.
Skill Vetting
A comprehensive security scanner designed to detect malicious code and suspicious patterns in AI agent skills before deployment.
XHS Auto
A professional automation workflow for generating, drafting, and validating Xiaohongshu social media posts.
Bomb Dog Sniff
A security-first management tool that sniffs out malicious payloads in Openclaw Skills before they can harm your system.
SolPaw
A professional tool for autonomous agents to launch Solana tokens on Pump.fun while maintaining full on-chain creator ownership.
Featured*
