SENTŌ
Agents sent to fight your battles.
Self-improving AI agents that run 24/7, powered by your Claude Code subscription. No pay-per-token API keys. One command to set up.
see it in action
here's a few. the rest is whatever you can think up.
autoplays on loop. tap a tab to jump to any of them.
Your Claude Code plan is the agent.
Most agent frameworks are open source, but to actually run them you need a pay-per-token API key, so the invoice starts ticking the moment your agent wakes up. Sentō uses the same Claude subscription you already pay for, and routes everything through Claude Code, which means more efficient token use on top of a flat monthly bill.
Sentō is an open-source AI agent framework built on top of the Claude Code CLI. Other autonomous agents are free to install but still require a pay-as-you-go Anthropic API key, so 24/7 operation racks up per-token charges. Sentō signs in with your existing Claude Code session (the same Pro or Max subscription you already have) and inherits Claude Code's context and caching optimizations, so every token goes further. The result is a self-improving agent on Discord, Telegram, Slack, or iMessage, online around the clock, with zero extra runtime cost.
- × separate API billing
- × metered per message
- × surprise invoices
- → uses Claude Code directly
- → flat monthly cost
- → no metering, no surprises
No pay-per-token keys.
Sentō signs in with your Claude Code OAuth token, the same one you already use. Every message counts against your Pro or Max plan, not a metered API invoice.
Your plan, your ceiling.
Pro gives you a solid daily driver. Max unlocks 24/7 heavy lifting. Same code, just more room to run.
Every Claude Code upgrade flows through.
New models, new tools, new skills. Your agent gets them the moment Claude Code does.
Built to stay online.
No servers to babysit. No glue to write. Your agent just keeps going.
One-Command Setup
npx sentoagent init. Answer a few questions. Your agent is live in 5 minutes.
Self-Healing
Guardian + Watchdog watch each other. Agent stays up. Fix anything from Discord.
Agent-to-Agent
First framework with secure inter-agent messaging. HMAC signed, mutual consent.
Personality Onboarding
Your agent interviews you on first message and configures itself from your answers.
11 CLI Commands
init, status, config, channels, skills, logs, update, start/stop/restart, doctor, pair, agents.
Multi-Platform
Discord, Telegram, Slack, iMessage. Runs on VPS, Docker, or your Mac.
Three steps. That's it.
- 01
Run the command
npx sentoagent init. Pick your model, tokens, and channel. Takes a minute.
$ npx sentoagent init - 02
Agent comes online
Launches on Discord, Telegram, Slack, or iMessage. Guardian watches the process.
▸ sento ✓ online on discord - 03
First message, first interview
Agent asks who you are and what you want. Configures itself. Ready to work.
sento: hey, who am I talking to?
Never set up a VPS or Discord bot before?
Full setup guideFrom zero to live agent in one command.
The full sentoagent init flow, exactly what you'll see in your terminal. Scroll down to start; it replays as you go.
11 commands. Zero surprises.
Everything you need, named how you'd guess.
- $npx sentoagent initSet up a new agent
- $sento statusCheck agent health
- $sento configChange model, tokens, keys
- $sento channelsAdd/remove communication channels
- $sento skillsList, install, export skills
- $sento logsView agent output
- $sento updateUpdate Claude Code, plugins, CLI
- $sento start/stop/restartProcess control
- $sento doctorDiagnose and fix issues
- $sento pairPair with another agent
- $sento agentsList agent info and pairs
Five layers. One loop.
Messages flow in. Guardian keeps the whole thing on its feet.
Sane defaults. No surprises.
HMAC-SHA256 signed messages
Every inter-agent payload is signed. No spoofing.
Mutual consent pairing
Both sides confirm before two agents can talk.
Rate limiting
Built-in ceilings keep runaway loops cheap.
DMs blocked by default
Opt in per-user. Nothing leaks until you say so.
Zero tokens for monitoring
Guardian watches the process, not your API budget.
The questions you're about to ask.
Does Sentō need a pay-per-token API key?
No. Sentō authenticates using the same Claude Code OAuth token you already sign in with on your machine, and that token is tied to your Claude Pro or Max subscription, not to a pay-as-you-go Anthropic Console key. Every message your agent sends counts against your subscription, not a separate metered invoice. If you prefer to run on a pay-per-token API key instead, Sentō supports that too.How much does it cost to run a Sentō agent 24/7?
Zero extra on top of the Claude subscription you're already paying for. A Pro plan is enough for a daily-driver agent; Max unlocks heavier 24/7 workloads. Sentō also routes everything through Claude Code, which means you get Claude Code's own context and caching optimizations, so the same token goes further than it would through a raw API call. You never see a separate token bill.How long does setup take?
About five minutes. Run `npx sentoagent init`, answer a few questions about model, channel, and bot token, and the agent launches and comes online on Discord, Telegram, Slack, or iMessage.What makes Sentō self-healing?
Every agent runs under a Guardian + Watchdog pair. Guardian restarts the agent process if it crashes; Watchdog watches Guardian. If the agent goes down, it's back up before you notice, and you can fix anything right from Discord without SSH'ing in.Can two Sentō agents talk to each other?
Yes. Sentō is the first framework with built-in, secure agent-to-agent messaging. Every message is HMAC-SHA256 signed, pairing requires mutual consent on both sides, and rate limiting is on by default.Which messaging platforms are supported?
Discord, Telegram, Slack, and iMessage out of the box. You can run on a VPS, inside Docker, or directly on macOS.Is Sentō open source?
Yes, MIT licensed. The source lives at github.com/sentoagent/sento and the package at npmjs.com/package/sentoagent.
Your agent is one command away.
Open source. MIT licensed. No vendor lock-in.
npx sentoagent init