Skip to content

eve

Connect eve agents to xmcp tools with an MCP client connection, and call an eve agent from an xmcp tool.

For the complete documentation index, see llms.txt. Markdown variants of every page are available by appending .md to the URL.

eve agents can consume xmcp tools through an MCP client connection, and an xmcp tool can start an eve session and return its reply. The runnable example does both with one read-only tool and one agent.

Shell

Give an eve agent xmcp tools

Protect the xmcp server with a bearer token checked on every request, using isBearer below with apiKeyAuthMiddleware as in the Flue guide. Then add a connection under agent/connections/ with defineMcpClientConnection:

agent/connections/xmcp.ts

eve evaluates connection modules during eve build, so keep required secrets inside callbacks such as getToken. tools.allow limits what the agent can discover.

Expose an eve agent as an xmcp tool

Protect the eve channel with its own token. isBearer compares the header in constant time:

src/env.ts
agent/channels/eve.ts

Then start a session per tool call with eve/client and wait for its result:

src/agent-client.ts

Call it from an ask-agent tool with getRequestContext().signal, as in the Flue guide. Keep ask-agent out of tools.allow so the agent cannot call itself.

Errors and cancellation

Each call creates one session and never retries. Failed runs, upstream errors, and empty replies return one generic tool error. Cancelling the MCP request stops waiting for the reply; the eve run may still finish.

Tested versions

eve 1.0.0-rc.18. The example's README covers environment variables, the eve server, and its tests.

One framework to rule them all