MCP.Market

Orbit MCP server

by constellation-works·io.github.constellation-works/orbit·v0.19.0·14 stars

Orbit's local-first task and workflow registry for coding agents.

B75/100grade B
What users say
No reviews yet
Be the first
Safety scan
B75/100

full report

Adoption
Growing

14 stars909 downloads/wk

Reviews

Write one

Nobody has reviewed Orbit yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Orbit tools

No tool declarations could be read from the package source. They show once the server is installed.

Public scan report

scanner v0.1.10 · 2026-10-16 · same rubric, same numbers if you re-run it

2 medium
  • Code scan2 source files scanned15/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 0 days ago15/15
  • Maintainer identityregistry namespace matches repository owner7/10

Findings (2)

  • mediumnpm install lifecycle script presentinstall.script
  • mediumnpm install lifecycle script presentinstall.script
    package.json: … "bin/orbit.js" }, "scripts": { "postinstall": "node scripts/install-binary.js", "…
Overall 75/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How Orbit compares

Against every graded Productivity MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
75/100
higher than 40% of 1,299 Productivity servers
GitHub stars
14
more than 80% of 854 Productivity servers
Downloads a week
909
more than 91% of 427 Productivity servers with a package

Install Orbit in Claude Code, Cursor or VS Code

Runs npx -y @orbit-tools/cli on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add orbit -- npx -y @orbit-tools/cli
Add to Cursor

What the publisher says

From the Orbit repository's README, as published. We do not edit it. Read it on GitHub

Orbit — Your agent files the work. Orbit ships it. You review the pull request.

The Orbit dashboard (orbit web serve) — task backlog, live audit log, per-agent scoreboard.

Say what you want in the coding agent you already use. Over MCP it files a task with acceptance criteria, Orbit runs it in an isolated worktree with the task's file scope reserved and a gated pipeline, and a pull request comes back for you to review. Every tool call, provider exchange, and task transition lands in a joined audit record. Local-first, driving the provider CLI you already have.

You drive Claude Code, Codex, Grok Build, or Gemini CLI against real code, often in parallel. Agents make it easy to skip the disciplines that keep code maintainable, and six months later nobody can reconstruct why an agent wrote a given line. Orbit makes those disciplines cheap without making you the clerk: the agent files the task before it edits, every tool call goes into a structured audit log, parallel runs are sandboxed into worktrees with file-level locks, and your own design docs are retrievable by the agents doing the work.

What a session looks like

Two commands set Orbit up. After that you talk to your agent, and Orbit does the filing, isolation, and delivery.

$ orbit init                                 # global state in ~/.orbit
$ cd my-repo && orbit workspace init --mcp   # .orbit/ here; operator-authorized MCP server
                                             # registered with your installed agent CLIs

You:    The fsProfile lookup is undocumented. Get that fixed.
Agent:  Filing it as a task with acceptance criteria.
        orbit.task.add → ORB-1042 · proposed
        Filed. Approve it into the backlog and ship?
You:    Yes.
Agent:  orbit.task.update   → ORB-1042 · proposed → backlog
        orbit.workflow.ship → run jrun-… · file scope reserved · worktree isolated
        Shipping. The run is asynchronous; I'll check on it.
        orbit.workflow.run.show → plan · execute · review settled · PR opened · ORB-1042 · review
        Pull request open. The diff and the merge are yours.

$ orbit task update ORB-1042 --approve       # after you merge the PR: review → done

Illustrative session, not captured output. Tool names are real; identifiers are placeholders and arguments are omitted.

Three things to notice:

  • The agent asks before the work starts. orbit.task.add lands the task in proposed. Moving it to backlog is the approval, a separate governed write that accepts nothing but a note. No agent session can skip it.
  • The run stops at a pull request. orbit.workflow.ship is review-only: it opens or updates a PR and leaves the task in review with the PR unmerged. Orbit and GitHub stay separate: approving the task never merges the PR, and merging the PR never completes the task.
  • Every step is on the record. orbit task show ORB-1042 reconstructs the prompt, plan, execution trace, and review thread months later, and every commit carries its task ID.

Features

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Orbit: common questions

Is Orbit MCP server safe?
Mostly: it is graded B (75/100). Read the Orbit safety report
How do I install Orbit?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Orbit need an API key?
Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is Orbit maintained?
The last commit was 19 days ago (2026-10-16). The latest release is v0.19.0.
What can I use instead of Orbit?
Servers from other publishers that do the same job: Argent MCP server, CLEO MCP Server and three.ws AgenC MCP server.
How does Orbit compare with other Productivity MCP servers?
safety score 75/100 (higher than 40% of 1,299 Productivity servers); github stars 14 (more than 80% of 854 Productivity servers); downloads a week 909 (more than 91% of 427 Productivity servers with a package). Full comparison.

Alternatives to Orbit

Same job from other publishers: the closest match first, then the best rated.

  • Argent
    Drive iOS Simulators, Android emulators, TVs and Electron/web apps from your coding agent
    A
  • CLEO MCP Server
    Task management for AI coding agents — CQRS 2-gateway pattern, 200 ops across 10 domains
    B
  • three.ws AgenC
    Browse the AgenC on-chain task marketplace, query the agent registry, and link identities.
    A
  • loci
    Cross-session memory and hybrid RAG over your notes and docs. Local-first, with cited answers.
    A
  • Better Notion
    Markdown-first Notion for AI agents -- pages, databases, blocks, and comments in one call.
    B