Skip to content

API.md

Latest commit

 

History

History
61 lines (41 loc) · 1.33 KB

API.md

File metadata and controls

61 lines (41 loc) · 1.33 KB

API

The hub exposes a small OpenAPI-compatible surface. The assistant talks to the hub, and the hub proxies selected calls to a rootd agent.

Authentication

Every hub request must include the control token:

Authorization: Bearer <CTL_TOKEN>

The hub uses the corresponding agent token internally when forwarding requests to rootd.

List servers

GET /servers

Returns currently registered agents.

curl -sS "$HUB_URL/servers" \
  -H "Authorization: Bearer $CTL_TOKEN"

Execute command

POST /srv/exec?server=<server-name>

Example:

curl -sS "$HUB_URL/srv/exec?server=server-100" \
  -H "Authorization: Bearer $CTL_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"cmd":"systemctl status nginx --no-pager", "timeout": 20}'

Typical response fields:

  • ok — whether the command succeeded;
  • returncode — process exit code;
  • stdout — command stdout;
  • stderr — command stderr;
  • duration — execution time when available.

System information

GET /srv/system/info?server=<server-name>

Returns basic system metadata for the selected server.

OpenAPI schema

See ../public/openapi.json. When deploying your own hub, update the schema servers.url to your public hub URL before importing it into ChatGPT Actions.