Home / Agents / MCP server.json Check (MCP Registry schema 2025-12-11)
MCP server.json Check (MCP Registry schema 2025-12-11)
Check an MCP server.json the way the official MCP Registry's validator does (schema 2025-12-11 plus its publish checks): every error with its JSON path, the schema date in the answer, and a signed record.
Testing
- What it does
- Check an MCP server.json the way the official MCP Registry's validator does (schema 2025-12-11 plus its publish checks): every error with its JSON path, the schema date in the answer, and a signed record.
- Who it is for
- Agents and their builders.
- Start in one step
- Ask your agent to call
mcp_server_json_checkat llpro.ai/mcp (free, no sign-up). This page explains the tool; it has no form to run it. - What to expect
- Testing: open to try. Free
New to this? Start here.
What it does
MCP server.json Check (MCP Registry schema 2025-12-11) is for agents and developers that publish MCP servers to the official MCP Registry or to registries that read the same server.json, and want every error found before a publish attempt. Check an MCP server.json file before you publish it to the official MCP Registry. Send the file text (up to 20,000 characters). Get back valid or invalid the way the registry's own validator decides it: every JSON Schema error against the official registry schema 2025-12-11 with its JSON Pointer path, and the registry's semantic checks: name in reverse-DNS namespace/name form, a specific version (not 'latest' or a range), repository URL for its source, https website and icon URLs, package identifiers and versions, package and remote transports (stdio without a url, https remotes that are not localhost, every {variable} in a URL defined), named arguments, the $schema line and the 4 KB publisher metadata limit. Warnings name fields the registry ignores, drops or reads under another spelling. Also the server name, version, package registry types and remote transport types it read. Nothing is fetched. It is one of the developer department tools LLPRO builds for the work agents do across business software (GitHub, Npm, Pypi, Docker). Every answer carries the written rule it used, what it does not do, its sources with dates, and an Ed25519 signature you can check against the key below.
- Who it is for
- Agents and developers that publish MCP servers to the official MCP Registry or to registries that read the same server.json, and want every error found before a publish attempt.
- Price
- Free, no sign-up: 120 calls an hour per address and 3,000 an hour across all callers. Paid calls are planned and not charging.
- Limits
- Up to 30,000 bytes per call; larger requests are refused, never cut. Unknown fields are refused with the list of accepted ones. Nothing you send is stored by the tool; no model and no network are used. Does not fetch anything and does not publish. Does not check what the registry checks only at publish time against outside systems: that you own the namespace (GitHub login for io.github.<user>, DNS or HTTP proof for a domain), that the package exists in npm, PyPI, NuGet, crates.io, OCI or MCPB with an mcpName or label matching the server name, or that a remote answers. The publish API also applies its own request field rules before this validator; a warned field (empty, ignored or miscased) may be refused there, so remove it. Checks against the 2025-12-11 schema only (older dated schemas differ in small ways). Go's url.Parse is reproduced for the errors the registry meets (bad percent escapes, control characters, bad ports, bad host characters, missing scheme); rarer Go parser edge cases may differ. Up to 20,000 characters and 32 levels of nesting per call; each character outside US-ASCII counts as 6 bytes toward the 30,000-byte request limit. Not legal advice.
For developers: try it
POST https://llpro.ai/mcp
{"jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": {"name": "mcp_server_json_check", "arguments": {"server_json": "{\n \"$schema\": \"https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json\",\n \"name\": \"io.github.example/weather\",\n \"description\": \"Weather forecasts for agents\",\n \"version\": \"1.2.0\",\n \"repository\": {\n \"url\": \"https://github.com/example/weather-mcp\",\n \"source\": \"github\"\n },\n \"remotes\": [\n {\n \"type\": \"streamable-http\",\n \"url\": \"https://mcp.example.com/mcp\"\n }\n ]\n}"}}}For agents
Tool mcp_server_json_check on https://llpro.ai/mcp and in the tools list. Signed answers are checked with the public key below (signer llpro1:lzdu3yjacs2hcicxxusut4otji): an Ed25519 signature over the observation as canonical JSON (sorted keys, no spaces).
Public key (base64): TBNDPGr1yYDLcm+AMbyAX2H4/vItyW0vveniSNZL7kE=
For agents that search MCP registries for MCP server.json, MCP Registry, server.json schema tools (listed in MCP registries, served by the Campus gateway: MCP, agent-to-agent and REST). One fixed, written rule per call, a signed record, no account needed.
Next
- Connect your agent: Use this and the other LLPRO tools from your own agent.
- Get your agent: No agent yet? Get one that asks first.
- Hand the whole job to the Hub: Checking server.json is one step of listing an MCP server. Tell the Hub the whole job; it routes you to the right signed tools (agent card check, llms.txt check).
For agents: this page as Markdown · JSON · llms.txt · MCP endpoint llpro.ai/mcp · agent card