Skip to content
agentd
@
~
Product
Overview
What agentd is, and where to start
Why agentd
The problem, the principles, the fit
Use cases
Shapes people deploy, with examples
Security at a glance
The containment model for reviewers
All documentation →
Get started
Install
Binaries, image, npm, source
Quickstart
A first agent, running, in minutes
Your first workflow
Write, run, kill, resume
Build a coding agent
A pair-programming daemon you host
All documentation →
Guides
Connect model providers
OpenAI, Anthropic, Bedrock, local, decisions
Connect MCP servers
Declare, authenticate, verify
Author workflows
Triggers, data, fan-out, resilience
Deploy
Jobs, systemd, Kubernetes
Harden a deployment
A checklist with validated config
All documentation →
Concepts
Architecture
Two loops, one binary
Workflows: how runs work
Durable DAGs, and why
Models, providers and roles
Kinds, providers, roles, fallback
Security model
Threat model and guarantees
All documentation →
Reference
Configuration reference
Every key, default and rule
CLI, environment and exit codes
Flags, env vars, exit codes
All documentation →
Editor
Docs
loading the editor…