$ hollowhost ai-jobs create --repo you/my-agent --lang python --pm uv --token-type NONE --follow

✓ AI Job created (ID: a1b2c3d4)

✓ Validating completed in 8s.

$ hollowhost ai-jobs deploy a1b2c3d4 --follow

✓ Deploying completed in 2m12s.

$ hollowhost ai-jobs run a1b2c3d4 --follow

✓ Run completed successfully in 4s.

── Logs ──

Analysis complete: 42 insights generated

Deploy Python AI Agents from Your Terminal

Push your Python, LangChain, CrewAI, or custom agent to GitHub and run it as a scheduled AI Job. Bring your own API keys, manage secrets and cron, follow runs from the CLI, and skip Dockerfiles, YAML, and cloud setup.

Deploy an AI Job → Explore Developer Workflow

Install

One Command to Deploy Agent Code

Install the CLI, log in, and deploy your first Python AI Job from GitHub.

# Install (macOS Apple Silicon)
$ curl -L https://github.com/.../hollowhost_darwin_arm64.tar.gz | tar xz
$ sudo mv hollowhost /usr/local/bin/


# Authenticate
$ hollowhost login
Opening browser for authentication...
✓ Logged in successfully!


# You're ready
$ hollowhost ai-jobs list
        

Full Control for Code-Based AI Jobs

Everything you need to deploy, run, and schedule your own agent code from the terminal. Built with Go for instant startup and zero dependencies.

Create, deploy, and run AI Jobs
Import .env files with secret auto-detection
Follow builds and runs in real time
Manage projects, storage, and schedules
JSON output for CI/CD scripting
Shell autocompletion (zsh, bash, fish)

Developer Experience

Built for Python and LangChain Developers

A CLI-first workflow for developers who want to ship agent code as scheduled AI Jobs without becoming cloud operators.

CLI-First Workflow Core

Create, deploy, and run AI Jobs from your terminal. Follow builds in real time with --follow. Import .env files with an interactive secret picker. Shell autocompletion included.

Per-AI Job Security Roles Security

Every AI Job gets a unique execution role. Database access scoped per AI Job. Zero lateral movement between AI Jobs.

Cron Scheduling Automation

Standard cron expressions with managed scheduling. Run AI Jobs every minute, every hour, or once a week. Toggle on/off without losing config.

Bring Your Own Keys Security

Use any LLM, any API, any service — we don't lock you in. Bring your own OpenAI, Anthropic, Mistral, or any other key. Secrets are encrypted at rest and injected at runtime, never exposed in logs or images.

Real-Time Log Streaming Coming Soon

Live log streaming to your terminal and dashboard — coming soon. Today, full execution logs are available after each run via the CLI or the dashboard.

.env Import DX

Import your .env file from the CLI. An interactive picker auto-detects sensitive keys and lets you classify each variable as env var or secret. Values never displayed.

Your AI Agent, Your Code

Write a standard Python or TypeScript agent. HollowHost handles containerization, deployment, security, secrets injection, and scheduling as an AI Job.

Python (pip, uv) or TypeScript (npm)
Secrets auto-injected as environment variables
Deploy and run from the CLI or the dashboard
Runs serverless (up to 15 min timeout)
Logs streamed to your terminal with --follow

# my_agent.py — your code, our infra
import os
import requests

def main():
    api_key = os.getenv("OPENAI_API_KEY")
    webhook = os.getenv("SLACK_WEBHOOK")

    # Your AI logic here
    result = run_analysis(api_key)

    # Post results wherever you want
    requests.post(webhook, json={
        "text": f"Analysis complete: {result}"
    })

if __name__ == "__main__":
    main()
        

Why HollowHost

How We Compare

We focus on one developer workflow: deploying code-based AI agents as scheduled AI Jobs with zero infrastructure management.

Capability	HollowHost	Cloud DIY	Heroku	Railway
Git-push deploy	✓	—	✓	✓
Per-AI Job security isolation	✓	Manual	—	—
Serverless (pay-per-run)	✓	✓	—	—
Cron scheduling	✓	DIY	✓	✓
Managed secrets	✓	DIY	✓	✓
Zero config Dockerfile	✓	—	—	Nixpacks
Token-based metering	✓	—	—	—
Dedicated CLI	✓	—	✓	✓
BYOK (any LLM/API)	✓	✓	✓	✓

Pricing

Predictable Pricing, No Surprises

Token-based. When you run out, runs pause — no overage charges.

Free

forever

500 tokens

Get Started

Starter

$9/mo

or $7.50/mo billed annually

12,000 tokens/month · Up to 10 AI Jobs · Up to 10 mins timeout

Choose Starter

Pro Popular

$39/mo

or $32.50/mo billed annually

60,000 tokens/month · Up to 50 AI Jobs · Up to 15 mins timeout

Choose Pro

Enterprise

$149/mo

or $124.17/mo billed annually

300,000 tokens/month · Up to 250 AI Jobs · Up to 15 mins timeout

Choose Enterprise