MCP Validator
Score and harden MCP tool manifests before you ship or publish a server agents will rely on.
Overview
MCP Validator is a MCP server for the Ship phase that scores MCP tool definitions from 0–100 on naming, schemas, annotations, and uniqueness.
What is this MCP server?
- Scores MCP tool definitions on a 0–100 scale
- Checks tool names, JSON schemas, annotations, and uniqueness
- Hosted streamable-http remote at wildrunai.com/api/mcp
- Open-source lineage via github.com/wildrunai/wildrunai at version 1.3.0
- Scoring scale 0–100
- Server version 1.3.0
- Remote https://wildrunai.com/api/mcp
What problem does it solve?
Publishing an MCP server with sloppy tool schemas wastes agent tokens and breaks client compatibility after you have already marketed the integration.
Who is it for?
Indie MCP authors shipping a first remote or local server who want a fast structural review without writing a custom linter.
Skip if: Teams that only need runtime penetration testing or who do not maintain MCP tool manifests at all.
What do I get? / Deliverables
After connecting the validator MCP, you get actionable scoring feedback to refactor tools before agents and users depend on them.
- 0–100 quality score per validation run
- Issue signals on names, schemas, annotations, and duplicates
Recommended MCP Servers
Journey fit
How it compares
MCP schema and naming linter via MCP, not a full application security audit or hosting platform.
Common Questions / FAQ
Who is MCP Validator for?
Developers and solo builders authoring MCP servers who want an automated checklist on tool quality before distribution.
When should I use MCP Validator?
Use it during Ship review or late Build when your tool list is stable enough to score, and again in Operate when you add or rename tools.
How do I add MCP Validator to my agent?
Add the streamable-http remote https://wildrunai.com/api/mcp in your MCP client, then invoke validator tools against your MCP definitions from the connected session.