Hi!
I’m building AgentSpec, a lightweight x402 API that inspects MCP tools and returns a machine-readable readiness signal before an agent connects to or publishes a server.
It can surface:
- risky or overly broad tool permissions
- missing or unclear input schemas
- agent-readiness score
- marketplace metadata gaps
Would an optional readiness signal be useful alongside MCP Registry entries?
The goal would not be to block publishing or replace registry validation. It would be an advisory, machine-readable preflight signal that registries, clients, or maintainers could choose to consume.
I’m looking for honest feedback and would be happy to provide a live demo.
Demo: https://agentspec-one.vercel.app/
OpenAPI: https://agentspec-one.vercel.app/openapi.json
Hi!
I’m building AgentSpec, a lightweight x402 API that inspects MCP tools and returns a machine-readable readiness signal before an agent connects to or publishes a server.
It can surface:
Would an optional readiness signal be useful alongside MCP Registry entries?
The goal would not be to block publishing or replace registry validation. It would be an advisory, machine-readable preflight signal that registries, clients, or maintainers could choose to consume.
I’m looking for honest feedback and would be happy to provide a live demo.
Demo: https://agentspec-one.vercel.app/
OpenAPI: https://agentspec-one.vercel.app/openapi.json