EvalGate MCP Server | MCP Hunter

EvalGate

Create and run AI evaluations from your agent

Finished Rank 3, top three
Votes 0
Launched Week 43, 2026 (19 to 25 October 2026)
Developer tools MCP servers Data & APIs MCP servers

From the maker

Create and run AI evaluations from your agent, then inspect the evidence behind each result. EvalGate lets authorized agents create evaluations, add test cases, start runs, create trace records, and review results, failure explanations, and baseline lineage. Access is scoped to your organization, role, and permissions. Writes follow EvalGate's existing authorization, quota, evaluation-lock, and spending controls. Public discovery exposes read-only tool schemas; product operations require authentication. A separate public documentation MCP helps agents find setup and usage guidance.

How to connect

Streamable HTTP endpoint:

https://www.evalgate.com/api/mcp

What we found when we connected

Verified 2026-10-18 | 12 tools | responded in 393ms

Protocol revision
2026-07-28
Server identified itself as
evalgate 3.14.1
Tools exposed
12
Tool descriptions
present
tools/list round trip
393ms
Tested
2026-10-18 13:51 UTC

The 12 tools it exposed

Every name below came back from this server's own tools/list response on the date above. So did each description, parameter list and label like "read-only": they are the server describing itself. We did not call any of them.

  • project.inspect read-only

    Inspect the evaluations and project keys available in the authenticated organization.

    Parameters: projectKey

  • project.plan read-only

    Return a non-mutating next-step plan for connecting a repository, evaluation, or trace source.

    Parameters: target

  • evaluations.list read-only

    List evaluations with a stable cursor for the organization.

    Parameters: limit, cursor, status

  • runs.list read-only

    List evaluation runs with stable cursors and regression signals.

    Parameters: evaluationId (required), limit, cursor, status

  • run.explain read-only

    Explain a run using its status, failed-case metadata, and evidence-lineage summary.

    Parameters: runId (required)

Show the other 7 tools
  • baseline.lineage read-only

    Show the published baseline anchor and its evidence lineage.

    Parameters: evaluationId (required)

  • docs.search read-only

    Search the published OpenAPI operation catalog for relevant help.

    Parameters: query (required), limit

  • connector.health read-only

    Show the latest connector-readiness reports for the organization.

    Parameters: connectorKey

  • connections.list read-only

    List current workspace connection metadata, with source restrictions, freshness and bounded continuation. Does not reveal secrets or grant source access.

    Parameters: q, source, limit, cursor

  • connections.catalog read-only

    Inspect all connector definitions and exact installed setup paths. Unavailable blueprints and live verification are separate facts.

  • connections.roots read-only

    Page independent configured account roots using current administrator authority. Does not list secrets or imply verified connector availability.

    Parameters: q, limit, cursor

  • connections.graph read-only

    Read canonical connection references and currently visible exact resource revisions. No source access grant or execution.

    Parameters: id (required)

Anything above describes one moment on the date shown. We do not re-test on a schedule, so nothing here says what this server is doing now. We have not audited its code, its data handling or its permissions, and this is not a safety rating.

Comments

What readers said about this product. Comments are people's own opinions and are not part of anything we checked.

Sign in to comment. Same account that votes.

Comments are written by readers, not by us, and we do not endorse them. Nothing here is held back or approved first: a comment goes up when it is posted, and you can edit or delete your own at any time. Report anything illegal or abusive with the link on each comment, or by email from our contact page if you do not have an account.