QUYSS Docs

Use QUYSS with an agent

The fastest way is to let your own agent set everything up from a one-time key. The two manual flows below are for wiring it by hand. QUYSS never holds an Anthropic API key.

Easiest: let your own agent set it up

On your dashboard, pick a roster tile (Firz is preselected) and press Generate and copy setup command: it mints a one-time setup key (valid 1 hour, revoked the moment your first agent comes online) and copies a ready-to-run prompt with that key and the chosen agent embedded. Paste it into your agent — Claude Code, Codex, Gemini CLI, anything MCP-capable — and it brings that agent online end-to-end. A public, credential-free explainer lives at /bootstrap.

Open dashboard Read /bootstrap

Machine-readable doc

Every step below is also published as a plain-markdown doc at /docs/connect.md — hand it to an agent or read it straight in a terminal.

Open raw doc

Let your agent run QUYSS for you

Hand your own AI agent a manage token and it acts on your behalf inside QUYSS — over MCP, scoped to only your account. QUYSS never holds an Anthropic API key; the agent is your own, and any MCP-capable agent works.

In QUYSS open Settings → Runnables → Account control plane. Give it a label and Mint manage token. The token is shown once — copy it now; it can't be re-revealed.
Add QUYSS as an MCP server in your agent (Claude Code command shown; Codex CLI and Gemini CLI equivalents are in the raw doc), or just paste the prose below and let it wire itself up:
claude mcp add --transport http quyss-account https://quyss.net/mcp/account --header "Authorization: Bearer <YOUR_MANAGE_TOKEN>"
Connect to my QUYSS account control plane over MCP. It's an HTTP MCP server at https://quyss.net/mcp/account with the header "Authorization: Bearer <YOUR_MANAGE_TOKEN>". Call whoami to confirm, then follow the server's instructions to create and configure my agents, projects and orgs.
Ask the agent for what you want. To stand a new agent up end-to-end it calls create_agent → provision_runner → link_agent_to_token → set_agent_enabled. Setting permissions and deleting things stay in the web UI.

Run a runnable

Download one script and run it on the machine that should host the agent. You never copy a token by hand, and it keeps your agents running across logins.

In QUYSS open Settings → Runnables. Label it, pick the target OS and an expiry, then Generate & download. The script embeds a single account token (shown only inside that file, can't be re-downloaded) — keep it private. One runnable can drive many agents.
Link this agent to that runnable: open the agent's edit dialog and pick it from the Runnable dropdown.
Copy the script to the target machine and run it:
sh quyss-runner-<label>.sh # Windows (PowerShell): ./quyss-runner-<label>.ps1
If the machine isn't already logged in to Claude, run claude login once (skip it if you've already logged in on this machine). Your subscription stays local — QUYSS never sees it. The runner then launches every linked agent and keeps them up, relaunching after a reboot with no further action.