Self-Hosted AI Agent Control Panel

AI Agents

A web dashboard on your own computer: assign tasks to AI agents, read the log of each run and track usage. It is not an agent itself and includes no model, and the data is kept in a local SQLite database. You bring it up on your own machine with Docker Compose, and the sign-in account and the API key are generated at random by a script. Derived from a frozen snapshot of an open-source project under the MIT licence; the delivery adds a single-service orchestration file, a Dockerfile that builds the image from source, a script that generates the account and keys, a demo script and a quick-start guide in Traditional Chinese.

Subscribe & download See how to start

SubscriptionUS$49 per month

Last run2026-11-02

In practice

Readiness4 / 5

What you need

Few changes needed4 / 5

Who it is for

Individual developers and small teams who want to assign tasks to AI agents and see run logs and usage in one place on their own machine.

How to start

$ sh scripts/quickstart/init-env.sh
$ docker compose up -d

Browse the files

Files in the product repository at the pinned commit. A few house files are left out of the zip.

File overview816 paths

▸docs24

SCREENSHOT-GUIDE.md
SECURITY-HARDENING.md
agent-setup.md
cli-agent-control.md
cli-integration.md
deployment.md
mission-control-activity.png
mission-control-agents.png
mission-control-cron.png
mission-control-memory.png
…

▸examples2

MULTI-PROVIDER-DEMO.md
first-task.sh

▸messages11

ar.json
de.json
en.json
es.json
fr.json
ja.json
ko.json
pt.json
ru.json
zh-tw.json
…

▸ops3

mc-provisioner-daemon.js
provisioner-limits.cjs
templates/

▸public34

brand/
llms.txt
mc-logo.png
mc.png
office-sprites/
robots.txt

▸scripts30

agent-heartbeat.sh
api-contract-parity.ignore
check-api-contract-parity.mjs
check-node-version.mjs
check-standalone-artifact.mjs
check-workflow-action-pins.py
deploy-standalone.sh
e2e-openclaw/
gen-star-history.py
generate-env.sh
…

▸skills4

mission-control-installer/
mission-control-manage/

▸src569

app/
components/
i18n/
index.ts
lib/
proxy.test.ts
proxy.ts
store/
test/
types/

▸tests79

README.md
actor-identity-hardening.spec.ts
agent-api-keys.spec.ts
agent-attribution.spec.ts
agent-comms.spec.ts
agent-costs.spec.ts
agent-diagnostics.spec.ts
agent-evals.spec.ts
agent-model-config.spec.ts
agent-optimize.spec.ts
…

▸wiki2

Home.md
STYLE_GUIDE.md

▫ATTRIBUTION.md

▫CHANGELOG.md

▫CLAUDE.md

▫CODE_OF_CONDUCT.md

▫CONTRIBUTING.md

▫Dockerfile

▫LICENSE

▫LICENSE-99LL.md

▫PRODUCT.md

▫QUICKSTART.md

▫README.md

▫RELEASE.md

▫SECURITY.md

▫SECURITY.upstream.md

▫SKILL.md

▫SUPPORT.md

▫delivery-manifest.txt

▫docker-compose.hardened.yml

▫docker-compose.yml

▫docker-entrypoint.sh

▫eslint.config.mjs

▫install.ps1

▫install.sh

▫next.config.js

▫openapi.json

▫openclaw_hardening_guide.md

▫package.json

▫playwright.config.ts

▫playwright.openclaw.gateway.config.ts

▫playwright.openclaw.local.config.ts

▫pnpm-lock.yaml

▫pnpm-workspace.yaml

▫postcss.config.js

▫tsconfig.json

▫vitest.config.ts

▫Config and hidden files 23

Latest passing run

Clone → Checkout → Toolchain → Build images → Control panel up and login

Pre-purchase Disclosure

What you are buying

A self-hosted AI agent task control panel: assign tasks to AI agents, read the log of each run and track usage. It is not an agent itself and includes no model. After running the account generation script you bring it up with Docker Compose.

Derived from an open-source project under the MIT licence, and not the original project itself. Our work: a Dockerfile that builds the image from source, an orchestration file, a script that generates the account and keys, a demo script and a quick-start guide in Traditional Chinese.

Delivery is a frozen snapshot: one zip, taken from the product repository at revision 5181258.

Known limitations

The service is bound to port 3000 of the local machine only; TLS and a reverse proxy are yours to bring
It is not an agent itself and includes no model
With no agent connected, the dashboard holds only the tasks you create and the sample data
No agent gateway is connected by default; connecting an OpenClaw gateway means running it yourself and editing and rebuilding the orchestration file
Connecting tools such as Claude Code or Codex needs your own accounts and keys
The upstream project itself is marked alpha, and the table structure and settings may change between versions
Google sign-in and the 1Password integration are not set up, and sign-in uses the local account and password
When the dashboard is open it queries GitHub for whether upstream has a new version number
The dashboard can run commands (its API includes spawn, pty and similar routes)
The data lives in a Docker named volume, and docker compose down -v deletes all tasks, agents and accounts
The first start builds the image from source, fetching from registries, apt and package repositories during that time
Pinned to the delivered snapshot; it does not follow upstream.

The liability cap, the scope of what is promised and the statutory floor are governed by the Terms of Service.

Vetting record

Everything below was checked against the pinned commit e28edf8. Every verdict carries its full reasoning, an evidence link and the checker; failures are published, not hidden — do not take our word for it, verify it yourself.

5 PASS0 WARN0 FAIL

Source ReviewPASS

Execution surface of the delivery reviewed item by item. Our own additions were read: docker-compose.yml (the published port is bound to 127.0.0.1 only; it refuses to start when the account, the password, the signing key or the API key is missing; only local host names are allowed; the data sits in a named volume; the container is read-only with the extra privileges removed and memory and process counts limited; no agent gateway is connected by default), scripts/quickstart/init-env.sh (openssl generates random values, .env is mode 600, and an existing one is not changed) and examples/first-task.sh (takes only the API key and port from .env and calls the local API). The upstream program can run commands (its API includes spawn, pty and similar routes); the Next.js and back-end sources were not read file by file, and the basis is the MIT licence, the local-only binding and the container limits.

Script SafetyPASS

The scripts on the buyer’s delivery path were each inspected: init-env.sh and the demo script are ours (read); the build goes through the Dockerfile (read: pnpm installs from the lock file, system packages and build tools come from apt, with no download piped straight into a shell); the image’s entry script is upstream’s and generates and keeps any missing secrets in the data volume. The run-stage image contains git, python3 and build tools for upstream’s agent installer feature.

RecencyPASS

This product is a frozen snapshot and does not follow upstream: the upstream source is pinned to the main-line revision taken on the day the product repository was created (see the evidence link), and the product repository is that snapshot plus our own commits.

LicensePASS

SPDX:LicenseRef-99LL-Product

Upstream LICENSE was checked word for word (the evidence link points at the file itself at the pinned version and matches the copy inside the product repository): the standard MIT text, copyright holder Builderz Labs. It allows commercial use, modification, derivative works, redistribution and sale; the duty is to keep the copyright notice and licence text (LICENSE and ATTRIBUTION.md ship in the zip); upstream has no NOTICE file. The decision does not rest on GitHub’s licence field. For the product as a whole the 99LL product licence template is attached: upstream’s LICENSE is kept as it is and LICENSE-99LL.md is added as the sale and liability terms for the part we built, with spdx recorded as LicenseRef-99LL-Product.

Dependency LicensesPASS

This records the declared side, not a scan result: the dependencies are declared in package.json and locked by pnpm-lock.yaml, and the packages themselves are fetched from their own repositories when the buyer builds; the base image comes from a registry and the system packages from apt. These dependency trees were not scanned package by package for licences. Upstream has no submodule and no third-party dataset was copied in (see ATTRIBUTION.md).