Agent Ops

_{Watch the 30-second video}

A team of AI agents and one human working one queue. Tasks, comments, reviews, delegation and decisions are Markdown files in a git repository, so every machine and every agent sees the same queue, and the history is plain text you can read and diff.

ops, a CLI that agents and people use the same way: create, claim, start, review, ask, answer, vote, write learn notes.
opsd, a background service per machine: it wakes agents (Claude Code, Codex, or any CLI you add) for work that is ready, keeps them within limits, watches their runs, and syncs the data.
The desktop app: inbox, board, task threads, agents, runs, schedules, channels, votes, knowledge and search, plus a tray icon to control opsd.

Status: beta (0.2.0-beta.1). Windows is the main platform today; the CLI and opsd also run on macOS and Linux from source.

Documentation: dawid-ai.github.io/agent-ops, built from this README and docs/.

A look around



Chat with the PM. Describe an idea; the PM splits it into tasks and routes them.	Every task is a thread. Claims, questions, answers, reviews and verdicts in one place.

Runs. What each agent is doing right now, its log, tokens and time.	Inbox. Only what needs you: questions, reviews and proposed tasks.

Install on Windows

Download AgentOps-Setup.exe and run it. It installs, for your user only:

the desktop app, with opsd inside it;
the ops command, added to your PATH if you leave the box checked.

The app updates itself from GitHub Releases: a new version downloads in the background and installs once no agent run is live and the window is closed. The installer is not signed yet, so Windows SmartScreen asks first: choose More info, then Run anyway.

On first start the app asks for your id (such as human:you) and the folder that holds your tasks, then starts opsd. To make a folder hold tasks, run ops init in it (see below).

To run agents you need their CLIs installed and signed in, such as Claude Code (claude) or Codex (codex). ops doctor checks them.

Run from source

Requirements: Node.js 22 or later, git.

npm install
npm run build
npm run desktop     # builds the UI and opens the app
npm run ops -- help # the CLI

npm run dist builds the Windows installer into packages/desktop/release/dist/.

Quick start with the CLI

The commands work the same in PowerShell, Command Prompt and bash.

Make a folder for your tasks. It must be a git repository: ops sync commits and pushes it, and other machines see the tasks once it has a remote. ops init does not run git init for you.
```
mkdir demo
cd demo
git init
ops init --name demo --human human:you --with-agents
```
--with-agents copies the starter role files into .agent-ops/agents/. ops agents lists them; only agents with active status can work, and opsd runs only agents that have a role file.
Register the folder and check the machine. If the desktop app's first start already chose this folder, skip ops setup; it does not add folders to an existing user config (the app's Settings does).
```
ops setup --me human:you   # writes the user config with this folder as a data root
ops doctor                 # user config, roots, claude and codex signed in, opsd
```

Walk one task through by hand. --as sets who acts; agents that opsd starts get their id in OPS_AGENT instead.

ops new "Write the parser" --to codex:backend -p p1 -m "Parse task files." --as human:you
ops next --as codex:backend                # T-0001  p1  Write the parser
ops start T-0001 "Plan: ..." --as codex:backend
ops progress T-0001 "Parser reads frontmatter" --as codex:backend
ops review T-0001 "Ready" --as codex:backend   # no --to: the human reviews
ops inbox --as human:you                   # review  demo/T-0001  from codex:backend
ops verdict T-0001 approve "Looks good" --as human:you

Start your first agent run. opsd starts paused, so resume it, then give an agent a task:

ops new "Add a README to this folder" --to codex:backend --as human:you
ops daemon start                           # the desktop app starts opsd for you
ops daemon resume
ops launch codex:backend --task T-0002 --budget 30m
ops runs                                   # the live run; ops logs RUN shows its output

The agent finishes with ops review, which lands in ops inbox. To have opsd start agents on its own when work arrives, set wake: true for them in the user config (see docs/supervisor.md).

ops help lists every command. The protocol, statuses and permissions are in docs/protocol.md; opsd is in docs/supervisor.md and the app in docs/desktop.md.

How data is stored

.agent-ops/
  config.yml                            # root name, id prefix, human, sync policy, leases
  agents/<runtime>-<role>.md            # role files: instructions, knowledge, style, limits
  tasks/<id>/task.md                    # frontmatter (title, type, owner, priority) + body
  tasks/<id>/events/<ulid>-<actor>.md   # one file per change, never edited
  kb/, channels/, decisions/, people/, projects/
  .ops/                                 # local only: index cache, locks, read markers

Writers only add new files, so git merges changes from several machines without conflicts. The current state of a task is folded from its events in ULID order, which every machine computes the same way.

Configuration

Root config (.agent-ops/config.yml, committed): name, id_prefix, human, sync (policy: manual, checkpoint or coordinated), leases, project (the project of a task that names none; default name), and project_aliases (old project names and the name each reads as now, such as v2: agent-ops).
User config (not committed): %APPDATA%\agent-ops\config.yml on Windows, ~/.config/agent-ops/config.yml elsewhere, or the path in OPS_CONFIG. Keys: me, roots, machine, id_block, and opsd (limits, providers, wakes, schedules, notifications). The app's Settings edits it.
Environment: OPS_AGENT (acting id), OPS_ROOTS (extra roots, separated like PATH), OPS_ROOT (root name for new tasks).

Development

npm run build          # tsc -b
npm test               # vitest
npm run lint           # eslint and prettier --check
npm run typecheck
npm run smoke:desktop  # renders every desktop view offscreen against a scratch root
npm run leaks          # scans for tokens, keys and webhook URLs
npm run docs           # builds the docs site into dist/docs and checks its links

Changes that users notice add a fragment to changelog.d/. See CONTRIBUTING.md and RELEASING.md.

License

MIT

#Agent Ops

#A look around

#Install on Windows

#Run from source

#Quick start with the CLI

#How data is stored

#Configuration

#Development

#License