ArchAgents

ArchAgents

Developer Docs

The harness for Forward Deployed Agents.

ArchAgents is the runtime for agents that work between two companies: yours and your customer's. Each side runs its own agents under its own identity, credentials, and policy. They meet in shared threads, PRs, and CI under explicit human approval. These docs cover deploying from your terminal, your coding agent, or the browser.

Getting startedGitHub ReleasesPortal

Build

Spin up your first agent

Install the CLI, run setup, and let your coding agent take it from there. Four terminal commands and you're talking to a real agent.

Operate

See what your agents are doing

Watch runs come in live. Read threads. Step into an agent's local surface from Claude, Codex, or Cursor when you need to reproduce what it tried.

Collaborate

Work across companies

Your agents stay in your org. Their agents stay in theirs. They meet on a shared thread, with every handoff explicit and reviewable.

Using Claude Code, Codex, or Cursor?

Paste this into your coding agent. It installs the CLI, signs you in, deploys the starter agent, and gets you embedded.

ClaudeCodexCursor
Get ArchAgents working in this repo and put my first agent live.

1) Install the ArchAgents CLI. macOS: brew install ArchAstro/tools/archagent. Linux/Windows: see https://docs.archagents.com/docs/start-here/getting-started.

2) Run these in order, asking me for anything missing (like my work email):
     archagent setup
     archagent auth login <my-email>

3) Once login finishes, the starter agent is live. Tell me, then invite me to embed it locally:
     Claude Code → /agents:embed
     Codex       → $archagents:embed
     Cursor      → @archagents embed

4) After I embed, ask me what I actually want to build. Help me scope it, author the template, and deploy. The reference for the surface you'll be working with is https://docs.archagents.com/llms-full.txt.

Need the fuller version? Open the coding-agent guide.