I originally contributed this as Core PR #228 (a health/status HTTP endpoint). @timur-ashkenov reviewed it there and directed me to follow the plugin creation instructions and submit it through the publication process:
Please follow the instructions for creating a plugin (https://github.com/singnet/Omega/blob/main/docs/reference-plugin-api.md), then submit a request in the "Issues" section according to the publication process (https://github.com/singnet/Omega/blob/main/docs/reference-plugins-publishing.md). The Omega team will publish your plugin on the wiki page (https://github.com/singnet/Omega/wiki/Omega-Plugins-Spreadsheet) once they receive your request.
I've reworked it into a standalone plugin accordingly, so I'm opening this request to have it added to the plugins list.
Plugin Naming
health — OmegaClaw health/status endpoint
Plugin URL
https://github.com/Leul-Negash/omegaclaw-health-plugin
Plugin Description
A MeTTa plugin that serves an OmegaClaw agent's liveness and high-level state over HTTP, so an orchestrator can poll it directly instead of scraping docker logs.
GET /health — 200 while the agent is alive, 503 once the heartbeat goes stale.
GET /status — JSON with status (thinking/sleeping/idle/offline), current_iteration, last_activity_at, uptime_seconds and version.
It reports state through the heartbeat API once per loop iteration, so the core loop needs no changes; the server, routes and staleness logic all live in the plugin. The port and reported version are configurable through config_get_by_key, so they can be set on the command line, through an OMEGACLAW_-prefixed environment variable, or in the config file, like the channel plugins.
It's useful for anyone running OmegaClaw under an orchestrator (Docker or Kubernetes liveness/readiness checks) that needs a real health signal rather than log scraping. The repository includes the plugin, a README with install and configuration steps, and a standalone self-test.
I originally contributed this as Core PR #228 (a health/status HTTP endpoint). @timur-ashkenov reviewed it there and directed me to follow the plugin creation instructions and submit it through the publication process:
I've reworked it into a standalone plugin accordingly, so I'm opening this request to have it added to the plugins list.
Plugin Naming
health — OmegaClaw health/status endpoint
Plugin URL
https://github.com/Leul-Negash/omegaclaw-health-plugin
Plugin Description
A MeTTa plugin that serves an OmegaClaw agent's liveness and high-level state over HTTP, so an orchestrator can poll it directly instead of scraping
docker logs.GET /health—200while the agent is alive,503once the heartbeat goes stale.GET /status— JSON withstatus(thinking/sleeping/idle/offline),current_iteration,last_activity_at,uptime_secondsandversion.It reports state through the heartbeat API once per loop iteration, so the core loop needs no changes; the server, routes and staleness logic all live in the plugin. The port and reported version are configurable through
config_get_by_key, so they can be set on the command line, through anOMEGACLAW_-prefixed environment variable, or in the config file, like the channel plugins.It's useful for anyone running OmegaClaw under an orchestrator (Docker or Kubernetes liveness/readiness checks) that needs a real health signal rather than log scraping. The repository includes the plugin, a README with install and configuration steps, and a standalone self-test.