mcp-protocol-conformance
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| runJsonRpcComplianceB | Run JSON-RPC 2.0 compliance suite (parse-error, invalid-request, method-not-found, invalid-params, internal-error) against a target MCP server. |
| runSpecVersionAssertionB | Drive an initialize handshake and assert the target advertises the expected MCP spec version. |
| runTransportSuiteC | Verify the target's transport layer (stdio line-delimited JSON or Streamable HTTP). |
| runOauthPkceFlowC | Exercise OAuth 2.1 PKCE S256 flow against the target. Mock-AS by default; real-tenant when --oauth-mode=real. |
| runToolSchemaValidationB | Validate that every advertised tool has a well-formed JSON-Schema inputSchema. Optionally diff against an expected manifest. |
| runCapabilityIntrospectionB | Compare the capabilities the target advertises in initialize against what tools/list, resources/list, prompts/list actually return. |
| runRoundtripSmokeB | Perform one tools/call per advertised tool. Fails on JSON-RPC transport errors, warns on tool-level isError=true. |
| runAnnotationsAuditC | Apply the heuristic annotation rules (destructive name regex, readOnlyHint conflicts) to every advertised tool. |
| runFullSuiteB | Run all (or selected) suites and produce a FullReport with summary + per-suite + per-check status. |
| generateReportC | Render a FullReport as JUnit XML, JSON, or human-readable terminal output. |
| compareManifestsB | Diff two tool manifests — added/removed/changed tools. |
| assertNoBreakingChangesB | Compare a baseline FullReport against a current FullReport. Returns a list of breaking changes and warnings. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 12 tools
Each tool targets a distinct conformance concern: JSON-RPC framing, spec version, transport, OAuth, schema shape, capability consistency, roundtrip, annotations, manifest diffing, and report rendering. runFullSuite intentionally aggregates the others but is clearly described as an orchestrator rather than a duplicate check. No two tools appear to do the same thing.
Nine tools share the runXSuite/runXAssertion camelCase verb_noun pattern, and the remaining three use equally descriptive distinct verbs (generateReport, compareManifests, assertNoBreakingChanges) that signal post-processing rather than test execution. The convention is predictable, though the run-prefix is applied to a few tools that are checks rather than suites.
Twelve tools is well within the ideal range and each maps to a concrete, non-redundant conformance check or reporting step. Nothing feels padded, and the granular suite tools plus a full-suite orchestrator is a natural decomposition.
The surface covers the major conformance axes: JSON-RPC, transport, spec version, OAuth PKCE, tool schemas, capability introspection, roundtrips, annotations, reporting, and regression detection. Gaps remain for server-initiated features (sampling, roots, elicitation), notification/progress behavior, and non-PKCE auth methods, but these are extensions rather than dead ends.