Your Autonomous AI Staff.
Deploy intelligent tentacles that work 24/7 — monitoring, researching, and acting on your behalf. Not an assistant you summon. A staff that never stops.
// Features
Deploy AI agents that run 24/7 in the background. They know their responsibilities and act independently.
A central intelligence that coordinates all tentacles, manages memory, and makes decisions across your staff.
Cron-like scheduling for tentacles. Set frequencies from minutes to weeks. Runs even when your computer is off.
Your tentacles don't wait to be asked. They monitor, detect changes, and report only what matters.
Write tentacles in any language — Python, TypeScript, or any executable. Standard IPC protocol for all.
Everything runs on your machine or VPS. Your data never leaves your infrastructure. Fully open source.
// Process
Clone the repo, set your API keys, and start the daemon. One command to boot your AI staff.
Install skill tentacles from the community or write your own. Each tentacle has a clear mission.
Tentacles run on their heartbeat schedule. They monitor, analyze, and act. Only important findings reach you.
// Use Cases
Each tentacle is a specialist. Deploy as many as you need.
Monitor competitor repos, products, and announcements 24/7. Get alerted only when something matters.
Scan communities for potential users, relevant discussions, and market signals. Draft outreach automatically.
Aggregate email, Slack, calendar, and news into one morning briefing. Delivered before you wake up.
Watch repos for important PRs, security issues, and dependency updates. Triage issues by priority.
24/7 monitoring of market anomalies, on-chain data, and trading signals. Acts on your strategy.
Run experiment loops, benchmark models, record results, and adjust strategy. Autonomous science.
// Architecture
┌──────────────────────────────────┐ │ Gateway Layer │ │ Feishu · Discord · Slack · API │ └──────────────┬───────────────────┘ │ ┌──────────────▼───────────────────┐ │ Brain Daemon │ │ ┌─────────┐ ┌────────────────┐ │ │ │ Memory │ │ LLM Gateway │ │ │ └─────────┘ └────────────────┘ │ │ ┌─────────┐ ┌────────────────┐ │ │ │ Scheduler│ │ Tool Registry │ │ │ └────┬────┘ └────────────────┘ │ └───────┼──────────────────────────┘ │ IPC (stdin/stdout) ┌─────▼─────┬─────────┬─────────┐ │ Tentacle │ Tentacle │ Tentacle │ │ hn-radar │ competitor│ briefing │ │ Python │ TS │ Python │ └──────────┴─────────┴──────────┘
The brain daemon is the always-on coordinator. It manages memory, schedules tentacles, routes messages through gateways, and makes cross-tentacle decisions.
Each tentacle is a subprocess with a clear mission. Written in any language. Communicates via standardized IPC protocol. Isolated, replaceable, composable.
Cron-like scheduler that wakes tentacles at their configured frequency. From every 5 minutes to once a week. Resilient to restarts.
Connect via Feishu, Discord, Slack, WhatsApp, or raw API. Multiple gateways can run simultaneously.
// Integrations
// Quick Start
Your first tentacle is one git clone away. Join the developers building their autonomous AI staff.