▄▀ ▀▄
▀██ ▄▄▄▄▄▄▄▄▄▄ ██▀
▀▄ ████████████ ▄▀
▀█████ ██████ █████▀
████████████
▀▄█████▄▀▀▀▀▄█████▄▀
▀████████▀
moltctrl
moltctrlShed Your Old
Agent Workflow
Security-hardened AI agent instance manager. Single binary, zero-config, process and Docker isolation.
// highlights
- ▸Single binary, zero dependencies. Download and run.
- ▸Interactive wizard or headless CLI flags for scripting
- ▸Process sandboxing with memory, CPU, and PID limits
- ▸Docker isolation with 15 security hardening measures
- ▸Multi-provider: Anthropic, OpenAI, Google, OpenRouter, AWS Bedrock, Ollama
- ▸Preconfigs for reusable agent templates across teams
// instance management
Create, start, stop, and destroy AI agent instances with simple commands.
moltctrl create --provider anthropic --name my-agent✓ Instance 'my-agent' createdmoltctrl start my-agent✓ Instance 'my-agent' started on port 3000moltctrl open my-agent✓ Opened http://localhost:3000 in browser// providers
One interface. Any model. Switch providers without changing your workflow.
| Provider | Env Variable |
|---|---|
| Anthropic | ANTHROPIC_API_KEY |
| OpenAI | OPENAI_API_KEY |
| GOOGLE_API_KEY | |
| OpenRouter | OPENROUTER_API_KEY |
| AWS Bedrock | AWS_ACCESS_KEY_ID |
| Ollama | OLLAMA_HOST |
// docker isolation
Run agents in hardened containers with 15 security measures including read-only root, no-new-privileges, and seccomp profiles.
moltctrl create --provider anthropic --docker --name secure-agent✓ Docker instance 'secure-agent' createdmoltctrl start secure-agent✓ Container started with security hardening// preconfigs
Save and reuse agent configurations. Share templates across teams and deployments.
moltctrl preconfig save my-setup --from my-agent✓ Preconfig 'my-setup' savedmoltctrl create --preconfig my-setup --name agent-2✓ Instance 'agent-2' created from preconfig ▄▀ ▀▄
▀██ ▄▄▄▄▄▄▄▄▄▄ ██▀
▀▄ ████████████ ▄▀
▀█████ ██████ █████▀
████████████
▀▄█████▄▀▀▀▀▄█████▄▀
▀████████▀Ready to molt?
Pinky is waiting. Download moltctrl and start managing your AI agents in seconds.