hivemind
Syncs the shared team plan through the repository's existing git remote using hidden refs (refs/hivemind/), where each member only writes to their own ref. The working tree, branches and PRs are never touched, and the local cache (.git/hivemind/state.json) is read in ~30 ms.
Primary tested git host for hivemind's hidden plan refs; the server detects GitHub public repositories and warns the agent and hivemind status that the plan is publicly readable.
Git host expected to work for hivemind's custom plan refs (accepts custom refs), though it has not been tested yet.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
hivemind
A shared, live plan for teams of coding agents.
You split the work, everyone's agents go off and build, and an hour later the plan in each agent's
head is stale: someone changed the User schema, someone else switched auth libraries, two agents are
building the same endpoint. hivemind keeps one plan in sync across every agent on the team and puts
teammates' changes into each agent's context at its next tool call.
npx hivemind-agents installRun it once per machine. No server, no account, no room codes: the team is the git repo.
What your agent sees
Teammate's agent changes an API while yours is mid-task. A few seconds later, after its next command, your agent sees:
[hivemind] Teammates updated the shared plan since you last checked:
ā ļø BREAKING alice-3f2a1c updated contract: `User.id`: uuid (breaking) (was: int)
šÆ (overlaps files you touched) bob-91d0e4 added task: [doing] Signup form @bob-91d0e4
- alice-3f2a1c added decision: JWT in an httpOnly cookie, not localStorage [inferred]
Check whether your current work depends on the breaking changes above and adapt it.At session start it gets the whole plan: goals, contracts (API shapes, schemas, shared types), decisions, who's doing what, and open questions.
Related MCP server: teamhub-mcp
Starting a plan
Just ask your agent, e.g. "We're 3 people building X this weekend. Make a plan for the team." It
records goals, contracts, decisions and tasks in hivemind (tasks unassigned, with dependencies), and
each teammate's agent can claim one by key. You can also add things yourself with hivemind add or
in the dashboard (hivemind open).
What belongs in the plan
The plan is for coordinating with teammates, not a log of each agent's work. Only what someone else needs to know or act on goes in: who is doing what, interfaces between people's work, team decisions, and questions for the team. An agent's own progress goes on its task. In a repo where only one person has ever worked, hivemind tracks just task status.
Plan and notebook
hivemind keeps two things apart:
The plan: goals, tasks, contracts, decisions, open questions. What the team agreed and is doing.
The notebook: status other people need, written on purpose (agents use
note_add; the extractor never writes notes), expiring after 7 days: ("the lab box is busy until 14:00", "precision 0.82 on the eval set", "image built locally, not pushed"). Agents see the last 24 hours of notes at session start and can search the rest withnotebook_get.
At session start agents see the most actionable things first (tasks in progress, then open tasks, contracts, decisions, questions, recent notes), with long sections capped so nothing important is cut off. Finished tasks and answered questions drop out of that view automatically.
Keeping the plan small
Plans that only grow stop being read. hivemind keeps them current on its own:
The extractor records less: contracts are only interfaces other people's work depends on (APIs, schemas, env vars, shared types), not component props or internal function signatures; decisions are only project-wide rules. It also sees recent commits, so finished work gets marked done.
A daily tidy (on plans with 40+ open entries, at most once a day per repo, by whoever's agent is active) closes what is superseded, duplicated or no longer true, marks tasks done that commits show finished, and moves status-like decisions to the notebook. Every change records its reason, and the run is logged in the notebook. Run it yourself with
hivemind tidy(dry run) /hivemind tidy --apply.Tasks get concluded: when an agent session ends, hivemind reads the end of its conversation and settles the tasks it worked on: done, todo (with what's left) or blocked. While a session runs, the extractor also checks its earlier tasks, so one it finished or moved on from doesn't stay "doing". A task that names a PR ("PR 123", "PR #123") is marked done as soon as that PR is merged into the main branch, whoever merged it and whichever session was told. Unfinished tasks and questions nobody touches for a week are closed as abandoned. A task stuck in progress shows "(no update for 6d)" until then.
When the plan changes direction
Plans change mid-hackathon. hivemind makes sure the old direction doesn't linger next to the new one:
New goals must say how they relate to existing ones: they replace them, or are a competing proposal. Competing goals are flagged to every agent ("the team hasn't picked one yet, ask the user") until someone resolves them.
Decisions and contracts can replace earlier ones, and agents are shown existing entries when they add new ones so they can spot contradictions. The background extractor removes entries that the work clearly abandoned.
hivemind review(or theplan_reviewtool) has the model check the whole plan for contradictions and suggest fixes (~15 s).
How it works
your agent āā hooks āāāŗ hivemind āāāŗ .git/hivemind/state.json (local cache, read in ~30 ms)
ā
background daemon: push your ref, fetch everyone's (every 3 s while active)
ā
origin: refs/hivemind/<member> (hidden refs, never branches)Sync through git. Each person's plan entries live on their own hidden ref (
refs/hivemind/<member>) on your existing remote. Nobody ever writes to someone else's ref, so there are no merge conflicts. Branches, PRs and your working tree are never touched.Updates arrive by themselves. Hooks add the full plan at session start, then only what changed since the agent last looked, with the most important items first.
The plan updates itself. When an agent finishes a turn, a background extractor reads only the new part of its transcript and pulls out the things teammates need: contracts, decisions and task progress, each with a quote as evidence. Agents can also publish explicitly through MCP tools (
contract_publish,decision_log,task_add,task_claim, ā¦), and they're told to for anything urgent.Work doesn't go unrecorded. Before an agent's first file edit, if it has no task in progress it's asked to claim one (and shown the unclaimed tasks). After a turn that changed files, if neither the agent nor the extractor recorded anything, it's reminded to update its task.
Updates arrive in seconds. Typically 2ā5 s from one agent's change to the other machine, then shown at that agent's next tool call.
Supported agents
Agent | Plan at start | Updates mid-task | Auto-extraction | Notes |
Claude Code | ā | ā every prompt and tool call | ā | |
Codex CLI | ā | ā every prompt and tool call | ā | Approve the hooks once with |
Cursor | ā | ā after each tool call | ā | Built from Cursor's hook docs; not yet tested in a live Cursor session |
install detects which of these you have and configures each one (--only=claude,codex to choose).
Claude Code plugin (alternative)
/plugin marketplace add Leonhest/hivemind
/plugin install hivemind@hivemindSame hooks and MCP tools, no npm needed. Use either the plugin or npx ⦠install, not both.
Commands
install puts a hivemind command on your PATH (or use npx hivemind-agents <command>).
hivemind status who's in the plan, last sync
hivemind plan print the shared plan
hivemind open live dashboard in your browser (and a form to add to the plan)
hivemind review check the plan for contradictions
hivemind tidy find finished, duplicate and outdated entries (--apply to clean up)
hivemind add <goal|task|contract|decision|question> <key> <text> [--breaking]
hivemind doctor check that everything is wired up
hivemind disable turn hivemind off for this clone (hivemind enable to undo)
hivemind uninstall remove hooks and MCP config (your repos' plan data is left alone)Privacy and cost
Your plan data never leaves your git host. There's no hivemind server. Anyone with read access to the repo can read its plan refs, the same people who can read the code.
Public repo = public plan. hivemind detects public GitHub repos and tells your agent (and
hivemind status) so nobody is surprised. Don't want that?hivemind disableturns it off for that clone;hivemind enableturns it back on.Secrets are scrubbed. Every plan entry is checked for keys, tokens, passwords and connection-string credentials before it's written, and the extractor is instructed to leave out credentials, personal details, customer names and unfixed security issues.
Extraction runs on your machine with your own agent login. It calls
claude -pwith Haiku (orcodex execif you only have Codex) using your existing subscription or API key: about $0.005 per extraction on API pricing, or a small slice of your subscription usage. SetHIVEMIND_EXTRACTOR_BACKEND=noneto turn it off and rely on the MCP tools only.Only new parts of transcripts with real work in them (a request or a code change) get extracted. Read-only exploration is skipped.
Requirements and limits
macOS or Linux, Node 20+, git, and a repo with an
originremote you can push to (without push access you still receive updates, but can't share yours).Tested with GitHub. Other hosts that accept custom refs should work; GitLab hasn't been tested yet.
Not inside a git repo, or no
origin? hivemind stays out of the way.
Troubleshooting
Run hivemind doctor. Logs are in .git/hivemind/ in each repo (hivemind.log, daemon.log, extract.log).
License
MIT
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelogā¦
This server cannot be deployed
Related MCP Connectors
Git collaboration for AI agents and people: tasks, hand-offs, reviews, teams, shared memory.
Shared control plane for AI coding agents ā tasks, memory, decisions, file locks. 12 tools.
The team layer for AI coding agents: shared contracts, collision alerts, E2EE sessions.
One shared project for a team's coding agents: decisions, a task board and handoffs.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceProvides MCP tools for AI coding agents to coordinate on shared repositories, enabling task claiming, conflict detection, and plan management in real-time.26 npmMIT
- AlicenseAqualityBmaintenanceEnables teams to share structured, git-synced context among AI coding agents working on the same repository, including living plans, task declarations, handoff briefs, file-provenance history, and conflict detection.129 npm2MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI coding agents on a team to coordinate by declaring plans, checking them against other in-flight plans and shared memory, receiving proceed/wait verdicts, and recording handoffs and rulings.MIT
- AlicenseNot gradedqualityCmaintenanceEnables multiple AI coding agents to coordinate on a shared software project by registering, claiming tasks, declaring file intents, publishing structured change reports, and handing off context, with a local dashboard showing state in near real time.MIT