Should a client declare the same capabilities on tools/list as on tools/call (2026-07-28)?

[buptliuhs]

Title: Should a client declare the same capabilities on tools/list as on tools/call (2026-07-28)?

Question

In 2026-07-28, client capabilities travel on every request in params._meta["io.modelcontextprotocol/clientCapabilities"] rather than once at initialize. That lets a server shape its responses per request, including tools/list.

But the spec doesn't seem to say anything about consistency between requests. In particular, can a client list tools while declaring one set of capabilities, then call those tools declaring a different set? If it can, a server has no reliable way to give the model a tool schema that matches how the call will actually be handled.

Is there a requirement or recommendation here, or should there be?

Why it matters

Take a server with a tool that needs user confirmation before it acts:

• The client declares elicitation: the server confirms through an elicitation dialog, so the tool's schema needs nothing extra.
• The client doesn't declare it: the server needs some other approach, for example an extra argument or different instructions in the tool description for the model to follow.

A server that does this returns a different tools/list depending on whether that request declares elicitation. That seems to be exactly the kind of per-request adaptation the envelope makes possible. It only works if what the client declares when listing matches what it declares when calling.

What we observed from one client (Claude.ai, 2026-11-13)

We connected the same MCP server from a Claude.ai Team plan account and a Free plan account, and logged what each request declared:

Request	clientInfo	Team plan	Free plan
tools/list on connect	Anthropic/Toolbox 1.0.0	{}	{}
tools/list from the chat	Anthropic/ClaudeAI 1.0.0	ui extension only	ui extension only
tools/call from the same chat	Anthropic/ClaudeAI 1.0.0	elicitation + ui	elicitation + ui
tools/list from "Refresh tools list"	claude-ai 0.1.0	elicitation + ui	ui extension only

Two inconsistencies stand out:

• Listing vs calling, on both plans. Every tools/list from connect or chat omits elicitation, while every tools/call declares it.
• Team vs Free, on refresh. The same "Refresh tools list" action declares elicitation on Team but not on Free. So a server that adapts its tool list shows different schemas on the two plans after a refresh, while the tool calls on both plans declare the same capabilities. The schema a user sees, and possibly the one the model works from, depends on the plan and on whether they've clicked refresh, not on how the calls are actually handled.

These are trimmed captures from the same connection (_meta only):

{"method":"tools/list","params":{"_meta":{
  "io.modelcontextprotocol/protocolVersion":"2026-07-28",
  "io.modelcontextprotocol/clientInfo":{"name":"Anthropic/ClaudeAI","version":"1.0.0"},
  "io.modelcontextprotocol/clientCapabilities":{"extensions":{"io.modelcontextprotocol/ui":{}}}}}}

{"method":"tools/call","params":{"_meta":{
  "io.modelcontextprotocol/protocolVersion":"2026-07-28",
  "io.modelcontextprotocol/clientInfo":{"name":"Anthropic/ClaudeAI","version":"1.0.0"},
  "io.modelcontextprotocol/clientCapabilities":{"elicitation":{"form":{},"url":{}},
    "extensions":{"io.modelcontextprotocol/ui":{}}}}}}

The result is that the model can hold a schema and instructions written for a client without elicitation, while every call it makes declares elicitation and is handled that way. A server that validates strictly against the schema for the capabilities declared on the call may then reject arguments the model was told to send.

Possible directions

1. Client guidance: within one connection, a client SHOULD declare the same capabilities on tools/list as on any tools/call that may follow, or at least a superset of what a later call will declare.
2. Server guidance: servers SHOULD NOT vary tool schemas by declared capabilities. They should keep one schema and branch only at call time. That's safe, but it gives up much of what per-request capabilities seem to enable.
3. An explicit statement that capabilities may legitimately differ per request, so servers know not to rely on tools/list.

Has anyone else hit this? Is there existing guidance we've missed?