A security-focused trust registry for verifying MCP servers and AI tools before installation.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install credence
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 credence 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 Credence?
Credence is a specialized security skill designed to protect AI environments by auditing MCP servers and AI tools against a community-verified trust registry. As one of the most vital Openclaw Skills for secure automation, it evaluates tools based on security, provenance, and behavioral risk. By integrating this registry into your workflow, you can prevent malicious or unstable tools from compromising your development environment, ensuring that every integration meets a high standard of safety and reliability.
Credence Use Cases
- Before installing any new MCP server or AI integration.
- When a user requests to add a tool with unknown security credentials.
- Verifying the safety and provenance of third-party repositories.
- Manual verification of tool trust scores using the /credence command.
How Credence Works
- The skill fetches the latest registry index from the Credence GitHub repository.
- It searches the registry for the specific server ID, name, or repository URL.
- It evaluates the trust score (0-100) and the corresponding verdict (e.g., APPROVED or REJECTED).
- If a tool is found, it reports the score breakdown and provenance details; if not found, it prompts the user to submit it for scanning.
- It provides a final recommendation on whether to proceed with the installation based on security thresholds.
Credence Setup
Ensure you have curl installed on your system. This skill is user-invocable and can be triggered via the command line or automatically during tool installation workflows. No API keys are required as the registry is public.
# Check a server manually via terminal
curl -s "https://raw.githubusercontent.com/pestafford/credence-registry/main/registry/index.json"
Credence Data Schema & Taxonomy
| Key | Description |
|---|---|
| server_id | Unique identifier for the MCP server |
| trust_score | Numerical risk rating from 0 to 100 |
| thinktank_verdict | Security status: APPROVED, CONDITIONAL, FLAGGED, REJECTED |
| attestation_file | Path to the detailed security breakdown file |
| repo_url | The source repository for the tool being scanned |
Credence Advanced Features
- Automated pre-installation scanning for all Openclaw Skills and MCP connections.
- Detailed attestation retrieval for deep-dive security audits including behavioral risk analysis.
- Support for partial match searching across server names and repository URLs.
- Seamless integration with adversarial AI deliberation for high-accuracy trust verdicts.
SKILL.md
Loading
Related Openclaw Skills
Docker Essentials
A comprehensive guide to essential Docker commands and workflows for container and image management.
Moral Compass
An ethical reasoning framework for AI agents grounded in universal principles like justice and truthfulness.
OpenKM REST Document Management
A Python-based CLI skill for automating OpenKM document management, metadata organization, and workflow execution via REST API.
OpenClaw Router
An intelligent model routing system that automatically selects the best LLM for any task to reduce costs by 60%.
Mission Control - ClawDash Pro
A professional skill designed to generate precise, design-preserving integration guides for connecting ClawDash Pro UIs to Open Cloud backends.
UseClick Link Shortening and Analytics
A powerful AI agent skill for automating UseClick link shortening, geo-targeting, and analytics workflows with plan-aware intelligence.
METADATA
Featured*
