Skip to main content
Glama

Server Details

Cloud console for a fleet of AI coding agents running on your own machines. OAuth.

Ownership verified
Status
Healthy
Uptime
100.0% over the last 24 hours
OAuth
Works in Glama
Last Tested
Transport
Streamable HTTP · MCP 2025-11-25
URL
Repository
agentxhq/agentx-mcp
GitHub Stars
0
Server Listing
AgentX MCP server
ChatGPT
–
Claude
–

TDQS

A4/5.0

Scored across 13 tools

Disambiguation4/5

Most tools target a clearly distinct resource+action (list_agents/get_agent, list_chats/list_chat_members/create_chat, list_artifacts/get_artifact, send_message/wait/interrupt_agent). The one genuinely confusable pair is `agentx_identity` vs `whoami`, which both read as 'who am I' and require reading the descriptions to separate. `check_agent_output` vs `wait` is also close but the descriptions draw a workable line (read transcript vs block for an event).

Naming Consistency4/5

The dominant pattern is consistent verb_noun (list_agents, get_agent, create_chat, send_message, interrupt_agent, check_agent_output), with matching prefixes for agent/chat/artifact families. Minor deviations: the stray `agentx_` prefix on identity, and the bare `wait` and `whoami` that break the verb_noun shape. Still fully readable and predictable.

Tool Count5/5

13 tools sits squarely in the well-scoped range and each one maps to a distinct capability: identity, agent discovery/detail/output/control, chat discovery/membership/creation, messaging/waiting, and artifact listing/reading. Nothing feels padded or redundant.

Completeness4/5

The core loop — discover agents and chats, send a message, wait or read the transcript, interrupt, inspect artifacts — is fully covered, and the server explicitly scopes agent creation to the web dashboard. Gaps are minor but real: no chat rename/delete or membership mutation, no artifact publish/write, and no way to list a chat's message history other than per-agent transcripts.

Publisher details

Operator
AgentX ApS
Operator website
https://www.agentxhq.com
Vendor relationship
First-party
Privacy policy
Not provided
Trust center
Unknown
Restrictions
Unknown

Available Tools

13 tools
agentx_identityA
Read-onlyIdempotent
Inspect

Show the user's own identity inside AgentX — the name their agents see when you message them on their behalf.

Inputs: no named parameters.

check_agent_outputA
Read-onlyIdempotent
Inspect

Read what an agent has been saying and doing — its recent transcript. Use this to catch up on an agent that is still working, or to re-read something it said earlier. Name the agent as list_agents shows it, and pass chat (from create_chat or list_chats) when reading by agent name. session_id targets one exact session and does not take chat.

Inputs: to, chat, tail, agent, +4 more.

create_chatA
Idempotent
Inspect

Create a new chat on a kick-to anchor agent. The server always mints the id and returns it as bubble_id — the chat's id, which other tools take as chat. Required: name and anchor_agent (as list_agents shows it). Does not touch connection grant allocation. Callers do not choose the chat id.

Inputs: name (required), anchor_agent (required).

get_agentA
Read-onlyIdempotent
Inspect

Show the full record for one agent — the detail list_agents leaves out, such as its model, capabilities, version and working directory. Name the agent as list_agents shows it. Rows do not show status: an agent can work in several chats at once, and a message to a busy agent is queued and answered.

Inputs: agent (required).

get_artifactA
Read-onlyIdempotent
Inspect

Read one file an agent has published. Pass uri as the mcp_uri field list_artifacts returned — form agentx://artifacts/{agent}/{path}. Large files are truncated and say so; use resources/read on the same uri for the full file, including binary.

Inputs: uri (required).

interrupt_agentA
Idempotent
Inspect

Stop an agent that is working right now, so it puts down what it is doing and listens. Use this when an agent has gone the wrong way and you want to redirect it — follow with send_message to give it new direction. Name the agent as list_agents shows it, and pass chat when stopping by agent name. session_id targets one exact session and does not take chat.

Inputs: to, chat, agent, session_id.

list_agentsA
Read-onlyIdempotent
Inspect

List the agents you can reach, a page at a time, each with the name to use when messaging it and the machine it runs on. Call this first — the names it returns are exactly what send_message, check_agent_output and interrupt_agent expect. People in the tenant are in the same list, marked kind: "human", and are messaged by name the same way.

Each row is a short summary; call get_agent with a name for the full record. Most recently active agents come first; the order may shift while agents work, so re-list from the start to refresh. When more remain, the result carries next_cursor — pass it as cursor for the next page. query matches the name, backend or state. An empty list means the user has no agents yet; agents are created in the AgentX web dashboard, not from here. Rows do not show status: an agent can work in several chats at once, and a message to a busy agent is queued and answered.

Inputs: query, cursor, page_size.

list_artifactsA
Read-onlyIdempotent
Inspect

List files the agents have produced and published — reports, diagrams, build output. Omit limit, or pass 0, for 100 files. Use next_cursor as cursor to continue, or query to filter by agent name or relative path. Name an agent to see only its files, or leave empty to see everything you can reach.

Inputs: to, agent, limit, query, +2 more.

list_chat_membersA
Read-onlyIdempotent
Inspect

List who is in a chat you name by id. A chat is the shared conversation — a session that holds other sessions. Pass chat (the bubble_id from create_chat or list_chats); the server never invents one and no connection has an implicit chat.

Returns the participants. It does not return message history — use check_agent_output to read what an agent said.

Inputs: chat (required).

list_chatsA
Read-onlyIdempotent
Inspect

Discover the tenant's chats and their ids. Each entry returns bubble_id — the chat's id, which send_message, list_chat_members, wait and the other chat tools take as chat. Page with limit (default 50) and pass next_cursor as cursor to continue; use query to filter by chat name or id. To see who is in one chat, use list_chat_members.

Inputs: limit, query, cursor.

send_messageA
Idempotent
Inspect

Send a message to one of the user's agents. Name the agent as list_agents shows it, and pass chat (from create_chat or list_chats) — that is the chat the message is delivered into. The server never invents a chat on send.

By default the message is handed over and this call returns. Set wait to true to hear the reply. session_id or address targets one exact session and does not take chat. Messaging a Core-homed human by name needs no chat either.

If the agent is still working when the wait ends, you get the latest part of its transcript and a note that it is still going — the agent keeps working, and check_agent_output will show how it finished. Never report silence as completion.

With wait true, status and next say how the call ended: replied / done means the reply is in wait.message or wait.output_snapshot, so don't wait again. Otherwise they describe the delivery: queued / wait means the message was handed over and no reply has been seen yet. On a snapshot, truncated means part of the output was left out: earlier messages when source is session_transcript (the latest message, and so a reply, is always whole), or the start of the text when it is output_view.

Inputs: to, chat, kind, name, +13 more.

waitA
Read-onlyIdempotent
Inspect

Pause until one of the user's agents does something in a chat you name — sends a message or finishes what it was doing — then return that event.

Pass chat (from create_chat or list_chats). The server never invents a chat for wait.

You rarely need this: set wait on send_message when you want the reply from an agent you just messaged. Reach for wait when you are expecting something from an agent you did not just message. Coming back after 110 seconds with nothing is normal and is not an error.

Inputs: chat (required), tail, lines, jitter, +3 more.

whoamiA
Read-onlyIdempotent
Inspect

Show who you are signed in as: the person, their organisation, and what this connection is allowed to do.

Inputs: no named parameters.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      B
      maintenance
      Self-hosted relay where coding-agent sessions on different machines share a team board, message each other and check for git conflicts before a commit or push. Supports isolated teams, private sessions, and an OAuth MCP endpoint for Claude in the cloud.
      9
      1
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables authorized AI coding agents to securely access development machines, repositories, terminals, and processes through a policy-enforced, audited control plane.
      Apache 2.0
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI coding agents to rent disposable cloud dev environments with a terminal, real browser, repo checkout, git gateway, and live view, so they can build, test, run dev servers, inspect pages, and push branches while keeping keys off the VM.
      1
      Apache 2.0
    Try in Browser

    Glama MCP Gateway

    Add one secure layer between your agents and this server.