AI makes purpose-built software easy to create. AgentCell makes it easy to deploy and share—with your partner, your team, or the five people who actually need it. No cloud expertise required.
An AI-native deployment platform for small web apps and internal tools. Works from Claude Code, Codex, Cursor, or any MCP client. Guide for agents →
No auth or infrastructure code added.
One control surface. Use it from
A tool for five people shouldn’t need a cloud built for five million.
Small software is purpose-built for you or your team: a workflow helper, sprint tool, prototype, tracker, or tiny app that solves one real problem.
Agents made these tools easy to build. AgentCell removes the cloud-shaped wall between “it works on my machine” and “my colleagues can use it.”
Your agent handles the infrastructure end to end. You decide who should use the tool; AgentCell takes care of everything between the folder and the front door.
No pipeline or runtime configuration. A folder of HTML or a frontend with a build script needs no Dockerfile: AgentCell detects it, builds it, and returns a live endpoint. A server brings its Dockerfile.
$ agentcell deploy Invite a person, a group, or your whole company. Colleagues sign in with their existing work account—no auth code or new account.
$ agentcell share team@company.com Every app runs in an isolated cell behind identity-aware access. Teams can bring approved packages, secrets, networks, and base environments.
$ agentcell access inspect Small tools may be used once a day or once a month. They sleep automatically and wake on demand, so keeping one around costs nothing.
idle compute · ₹0 / $0 Big clouds expose infrastructure and expect you to assemble a platform. AgentCell exposes outcomes. Everything your tool needs remains operable by your agent or from the CLI.
agentcell deployBuild and release
agentcell logsSearch and stream
agentcell metricsHealth and performance
agentcell envSecrets and config
agentcell rollbackInstant recovery
agentcell accessIdentity and sharing
Deploy every useful tool and half-promising experiment. AgentCell only charges when somebody actually opens one and compute runs.
Illustrative usage, not final pricing.
Short answers, honest about what ships today. Coding agents get the same text at index.md.
https://<cell>.agentcell.cloud, and the coding agent that wrote the app can deploy and operate it through a CLI or an MCP server. A container app or a static site: a folder of HTML or a frontend build is served straight from the platform's edge, with no machine running for it.agentcell login signs you in with Google, GitHub or an emailed PIN and creates the account. agentcell deploy --cell my-app . uploads the directory, builds it and prints the URL. agentcell logs my-app streams the output. The whole path is in Deploy an app in three commands.agentcell binary is also an MCP server: add {"command":"agentcell","args":["mcp"]} to the agent's MCP configuration and it gets deploy, logs and ps as tools, with the same permissions as the CLI. Decision criteria and per-harness setup are on AgentCell for AI coding agents.index.html, or a package.json with a build script (Vite, Create React App, Vue, Svelte, Astro, Next.js with output: 'export'), which the platform builds. Anything else that runs as one container deploys from a Dockerfile: Next.js in server mode, FastAPI, Flask, Django, Streamlit, Express, Go, Rails. A container listens on the port it EXPOSEs, 8080 if none, and /data persists across restarts. Working examples are in the samples repository.index.html, or a package.json with a build script, deploys without one; use client 0.1.4 or later, which leaves node_modules out of the upload. A server (FastAPI, Flask, Streamlit, Express, Next.js in server mode) still needs a Dockerfile: a coding agent writes one in seconds, and the samples repository has several to copy. Detecting servers without a Dockerfile is planned, not shipped.agentcell.cloud URL, with its own origin; holding the link is not enough. Public sites with no sign-in are not available. Letting named people or an organisation in is the sharing feature above, in private beta. Read The membrane for what the isolation covers./data survives restarts and redeploys. SQLite in /data is the natural fit for a small tool, and the notes-sqlite sample shows it. For Postgres or Redis, connect to a hosted service over the network; a cell is one container, with no sidecars. A static site has no /data.agentcell deploy uploads the directory in front of you; no repository, pipeline, YAML or AWS account is involved. Retrying is safe because the directory hash is the idempotency key. See deploy an app without DevOps.deploy, logs, ps, whoami. Listed by the CLI and MCP server but not yet shipped: rollback, env, secrets, domains, share, access, spend, destroy; each answers a typed not_found until it does. Sharing and the identity-aware front door are in private beta.agentcell login, and deploy a directory: three commands. Coding agents should read the agent guide or llms.txt. Anything else: hello@agentcell.dev.Bring the tool currently stuck on localhost. Sign in with the account you already have and deploy it in three commands. No cloud account, no YAML, nothing to configure.
$ agentcell login
✓ signed in as you@example.com
$ agentcell deploy --cell my-app .
✓ https://my-app.agentcell.cloud
$ agentcell logs my-app
listening on :8080