Skip to main content
Glama
nullbuilds-ai

null-trust-score

null-trust-score

MCP server that exposes trust scores for MCP servers. Query before you connect.

Built by NULL (@nullbuilds).

What it does

Any AI agent can query this MCP server to check whether an MCP server is safe to connect to. Returns a composite trust score (0-100), letter grade (A-F), per-dimension breakdown, and recommendation.

53 MCP servers scored. 3 with full security reviews (46 findings). 50 auto-scored from public signals.

Related MCP server: dominion-observatory

Tools

Tool

Description

trust_get_score

Get trust score for a specific MCP server

trust_search_servers

Search servers by name, score, or grade

trust_get_review

Get full security review with findings

trust_list_flagged

List servers graded D or F

trust_check_batch

Check multiple servers at once (max 25)

Scoring

6 dimensions, weighted composite:

Dimension

Weight

What it measures

Security

25%

Input validation, credential handling, sandboxing

Reliability

20%

Error handling, graceful degradation, timeouts

Documentation

15%

README quality, tool descriptions, examples

Maintenance

20%

Commit recency, release cadence, dependency freshness

Design

10%

Tool naming, response format, annotations

Ecosystem

10%

Downloads, stars, directory listings

Grades: A (80-100), B (60-79), C (40-59), D (20-39), F (0-19).

Servers with full reviews get scores across all 6 dimensions. Auto-scored servers (from public signals only) get a scoring_version: "auto-1.0" tag.

Installation

npm install
npm run build
npm run seed   # populate trust.db with 53 servers

Usage

With Claude Code

Add to your .mcp.json:

{
  "mcpServers": {
    "null-trust-score": {
      "command": "node",
      "args": ["/path/to/null-trust-score/dist/index.js"]
    }
  }
}

With MCP Inspector

npm run inspect

Data sources

  • Reviewed servers: Full source code audit by NULL. Findings verified against actual code.

  • Discovery queue: 50 top MCP servers ranked by downloads, stars, first-party status, and recency. Auto-scored using maintenance and ecosystem signals.

License

MIT

F
license - not found
-
quality - not tested
D
maintenance

Maintenance

Maintainers
Response time
Release cycle
Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • A
    license
    A
    quality
    F
    maintenance
    Enables AI agents to query trust scores for MCP servers and agent skills while scanning content for potential security issues. It provides direct access to the Vigile trust registry to help users evaluate the safety of third-party tools and integrations.
    Last updated
    9
    76
    MIT
  • A
    license
    -
    quality
    A
    maintenance
    Behavioral trust scoring for MCP servers and AI agents. Live registry tracking 4,500+ servers with trust scores (0-100) based on interaction history, success rates, and latency
    Last updated
    2
    MIT
  • A
    license
    -
    quality
    D
    maintenance
    MCP server for AI agent trust verification, enabling agents to verify identities, check trust scores, and build reputation across multiple blockchain and web platforms.
    Last updated
    22
    1
    MIT

View all related MCP servers

Related MCP Connectors

  • Pre-connect trust checks for AI agents and MCP servers using HVTracker's public trust registry.

  • Independent A-F trust grade for any MCP server, watched for drift. Free, never for sale.

  • MCP server teaching AI agents to implement TideCloak: auth, E2EE, IGA, security analysis

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/nullbuilds-ai/null-trust-score'

If you have feedback or need assistance with the MCP directory API, please join our Discord server