The full stack to ship agents in customer environments.
Gateway, control plane, and runtime — so every agent runs under each user's and each agent's own permissions, with hard budgets, an off switch, rollback, and a verifiable record of every action.
- Any model, cloud or local
- Any agent engine
- One UI for every agent
› The stack around your agent
Only lease heartbeats, budget counters, and audit metadata cross this line — never prompts, never data.
- LangChain
- CrewAI
- Agent Pack
- any SDK
- Postgres RLS
- Unity Catalog
- DynamoDB
- OpenTelemetry
- S3 Object Lock
› Run your first governed agent
On your laptop or a server — either way, you govern it from control.agenthippo.ai.
- Docker Desktop with WSL 2 + Ubuntu
- Docker Desktop
- Docker
- AgentHippo IDE
- An agent pack — any folder with
agent.yaml, or a demo pack
-
Turn on Docker for WSL once
Docker Desktop → Settings → General: tick Use the WSL 2 based engine. Then Settings → Resources → WSL Integration: switch on Ubuntu and click Apply & Restart.
-
Register your agent and download its bundle
Sign in at control.agenthippo.ai, click + register agent, and download the bundle. Extract it anywhere — for example
C:\Users\<you>\Downloads\<agent>-bundle\<agent>~/Downloads/<agent>-bundle/<agent>. -
Bootstrap the agent
In the Ubuntu terminalIn a terminal, point the bundle at your pack:
cd /mnt/c/Users/<you>/Downloads/<agent>-bundle/<agent> bash ./bootstrap.sh /mnt/c/path/to/your-packcd ~/Downloads/<agent>-bundle/<agent> bash ./bootstrap.sh ~/path/to/your-packIt worked if the output ends with:
✓ agent up on 127.0.0.1:8095✓ IDE: open stack/ide-pack-local in the AgentHippo IDE — the loopback edge on 127.0.0.1:9095 verifies your sign-in -
Chat with it in the IDE
Open the AgentHippo IDE and sign in. In the chat panel, click the agent picker at the top right (it shows Default) → Install Agent from Folder, and choose the bundle's
stack/ide-pack-localfolder. Pick your pack from the same menu and start a chat — it runs on your local agent. -
Control it from the console
Watch usage at control.agenthippo.ai. Kill the agent and send a message: it's rejected. Resume it and send again.
- Any VM with Docker
ssh+rsyncon your laptop- Ports 80/443 open for web access
- An agent pack — any folder with
agent.yaml, or a demo pack
-
Register your agent and download its bundle
Sign in at control.agenthippo.ai, click + register agent, and download
<name>-bundle.tar.gz. It's prewired to your account. -
Deploy from your laptop
Point the bundle at a pack and your VM. Signing happens automatically on the VM.
tar xzf <name>-bundle.tar.gz && cd <name> AGENT_VM=root@your-vm ./bootstrap.sh /path/to/agent-packs/agenthippo-emoji-firstYour agent appears in the console about 15 seconds later, signed and governed. No provider key ever exists on the VM.
-
Chat with it from the web
Once the console assigns a hostname, bootstrap turns on HTTPS and single sign-on automatically. Open the agent's address, sign in, and chat — every request carries the signed-in user.
The demo packs include the chat page; add a
www/index.htmlto your own pack for the same. -
Or from the IDE
Open the bundle's
stack/ide-pack/folder in the AgentHippo IDE.
› Then put it to the test
Cap the spend
Set a budget and watch token costs meter live. The call over the cap is refused, not flagged.
Hit Stop
Kill the agent from the console. It stops within one poll — no SSH — until you resume it.
Ship v2, then roll back
Run ./bootstrap.sh with a new pack folder, then roll back to the last version.
Pull the audit trail
Who ran what, which version, and what it cost — for every turn.
› No agent yet? Start with a demo pack
agenthippo-emoji-first Start here · ~2 min Zero-config starter. The fastest way to see the whole loop. View on GitHub quickstack-analyst Per-user data Two users ask the same question and each sees only their own rows. Needs Postgres with row-level security. View on GitHub Running quickstack-analyst? Ask us for the seed scripts.
› The IDE and CLI are free
AgentHippo IDE
Build packs and chat with your agents.
- Any model, cloud or local Anthropic OpenAI Gemini Bedrock Vertex local models
- Any agent engine Claude Code Codex OpenClaw your own
- One UI for every agent plus Slack WhatsApp Telegram
AgentHippo CLI
Serve, automate, and run in CI.
After download: unzip agenthippo.zip && ./install.sh
curl -fsSL https://agenthippo.ai/install.sh | bash AgentHippo IDE
Build packs and chat with your agents.
- Any model, cloud or local Anthropic OpenAI Gemini Bedrock Vertex local models
- Any agent engine Claude Code Codex OpenClaw your own
- One UI for every agent plus Slack WhatsApp Telegram
AgentHippo CLI
Serve, automate, and run in CI.
After download: Expand-Archive agenthippo.zip; .\agenthippo\install.ps1
irm https://agenthippo.ai/install.ps1 | iex AgentHippo IDE
Build packs and chat with your agents.
- Any model, cloud or local Anthropic OpenAI Gemini Bedrock Vertex local models
- Any agent engine Claude Code Codex OpenClaw your own
- One UI for every agent plus Slack WhatsApp Telegram
AgentHippo CLI
Serve, automate, and run in CI.
After download: unzip agenthippo.zip && ./install.sh
curl -fsSL https://agenthippo.ai/install.sh | bash
Not sure which? Run uname -m — aarch64 means ARM64.
AgentHippo IDE
Build packs and chat with your agents.
- Any model, cloud or local Anthropic OpenAI Gemini Bedrock Vertex local models
- Any agent engine Claude Code Codex OpenClaw your own
- One UI for every agent plus Slack WhatsApp Telegram
AgentHippo CLI
Serve, automate, and run in CI.
After download: unzip agenthippo.zip && ./install.sh
curl -fsSL https://agenthippo.ai/install.sh | bash
Not sure which? Run uname -m — aarch64 means ARM64.
All builds are signed. Checksums · Release manifest