Skip to content
Agents tracked: 502 Releases (7d): 762 Agent status: no reported issuesSince last limit reset: Claude 11d · Codex 9d For AI agents: Markdown Updated Nov 22, 2026

emisar

by andrewdryga

let your AI work on servers with approvals.

OtherOfficialRemote

GitHubWebsite

1. Made by

“portal/ Elixir/Phoenix control plane, operator console, website, and MCP API”

The repo holds emisar's own control plane, bridge and runner; its README sends users to sign up at emisar.dev, the site links back to this repo, and the registry name dev.emisar sits on that domain.

github.com/AndrewDryga/emisar/blob/31f4c55b3f7513621e09e3f4468ff29dfdc6e123/README.md

Checked against its repository at one commit, on Nov 10, 2026. Tap a verdict to see its source.How we checkedReport an error

What emisar does

emisar lets an AI assistant check and fix your servers without handing it full remote access. It can only pick from a list of prepared actions, like reading logs or restarting a service. Your rules decide what runs, what waits for a person to approve, and what is refused, and every step is logged. It needs an emisar account and a small program on each server.

What people use it for

Look into an incident on live servers

When something breaks, your assistant reads logs, service status and disk use across your servers and comes back with evidence and a likely cause.

Restarts that wait for your yes

Routine checks run on their own, while restarts and deployments pause until a person approves them in the emisar console.

Find and run a prepared action

The assistant searches the action list by task, reads the exact inputs one action takes, then runs it on the servers you name.

Check what your own deploy did

After a release misbehaves, the assistant looks at the running result, finds its own mistake and proposes a fix for you to review.

Turn a fix into a reviewed runbook

The assistant writes a repeatable list of steps as a draft; a person reviews and publishes it before anyone can run it.

See what the assistant did last night

Review every run the assistant started, with who asked, the reason, the policy decision and the result.

emisar popularity

449
GitHub stars, andrewdryga/emisar
0
public agent projects connect it
emisar: GitHub starsandrewdryga/emisar 449

Tools it gives your agent (14)

What it sends and its license

Sends

every action request with its exact arguments, policy decisions, approvals and the redacted output

To emisar control plane on Google Cloud in the United States; every tool call that runs or records an action. Opt out: none; mark sensitive action fields so they are masked in views (they stay in storage).

Sends

command output from your servers, filtered on the server before it leaves

To emisar run history; every run. Opt out: per-action redaction rules.

Sends

product usage events (not your servers' data or action contents)

To Mixpanel; while you use emisar. Opt out: none stated.

The server program, bridge and action packs are open source under Apache 2.0. The control plane is under the Business Source License, and the hosted service is free for 3 servers and paid per server above that.

In a paid app
With conditions

Use the hosted service under its plans (free for 3 runners and 1 user); running the control-plane code yourself in production needs a commercial license.

“Other production use requires a commercial license.” source

Pass on its files
With conditions

The runner, bridge, packs, skills and Cursor plugin are Apache 2.0; the rest of the code is Business Source License 1.1 until its change date.

“`runner/`, `mcp/`, `packs/`, `skills/`, and `dist/cursor-plugin/` are open source under the [Apache License 2.0](runner/LICENSE)” source

client-work
With conditions

Fine on the hosted service under a plan that fits the client's server count; each person's key acts with their own access.

“Free for 3 runners” source

Is it alive

Latest version

0.60.0 in the registry, updated Nov 18, 2026

Last release

0.58.0 in the MCP Registry (published 2026-11-08); code commit 31f4c55 of 2026-11-09

Remote endpoint
yes: an unauthenticated GET to https://emisar.dev/api/mcp/rpc on 2026-11-10 answered 405 (the endpoint takes POST; real calls need sign-in)

Alternatives

emisar This page

449 starsRemote

ansible-core

Pick it if you want repeatable server configuration written as reviewed playbooks

Local

mcp-server-commands

Pick it if the agent only needs to run commands on your own disposable machine

39.6K starsLocalin 1 projects

io-github-coder-coder

Pick it if the agent should manage cloud development workspaces

Local

All other servers

Get started

Tell your AI assistant

“Use the emisar server to find out why the api service on web-1 is failing; use read-only actions first, give a reason on every call, and ask me before any action that needs approval.”

Connect it

claude mcp add --transport http emisar https://emisar.dev/api/mcp/rpc

Claude.ai and ChatGPT connect to the remote address with OAuth sign-in; Claude Code, Codex, Cursor local mode and other local clients use the emisar-mcp bridge, whose connect command writes their configs. The console's AI agents page generates the exact per-client lines; the README gives no Claude Code or Codex command. The cursor line is dist/cursor-plugin/mcp.json.

Mistakes to avoid

Calling run_action again after its reply was lost
Recover it with get_operation and the same key; never repeat the change

A lost reply does not undo work the control plane already accepted.

Calling run_action straight from a find_actions result
Call get_action first for the exact inputs, risk and side effects, then run_action

The selected action may be destructive and its inputs are strict.

Running find_actions once per server
One find_actions query covers every server in scope

Repeated searches only waste calls.

Committing .mcp.json with EMISAR_API_KEY to the project
Keep the key in user-level or secret config; one key per client

The key acts as you on your servers.

Retrying the same call after invalid_args
Fix the exact paths the error names, then call again

The error lists what to correct; the same call fails again.

Questions

Can the assistant run any command it wants?

No. It chooses from declared actions with typed inputs; the server program rejects unknown actions and inputs.

Who approves risky actions?

A person, in the emisar console. Approvals are never decided inside the assistant's chat.

Does my server need an open port?

No. The program on the server only connects outward over an encrypted connection and listens for nothing.

Is my data used to train AI?

emisar says it does not use your data to train AI models.

Is it free?

Free for 3 servers and 1 user with no card; Team is $20 per server a month.