Scoreboard dashboard for fipsagents-based AI agents. Polls agent REST endpoints and displays trust level, maturation stage, learned skills, and work-item queue depth in a single-page HTML dashboard.
Not yet implemented. See docs/architecture.md for the design direction and the tracking issue for the implementation plan.
- Single-file HTML — no npm, no build step, no CDN dependencies. One self-contained
index.htmlwith inline CSS and vanilla JavaScript. - Standalone deployment — deploys as its own container and OpenShift route, following the same pattern as ui-template and gateway-template.
- API consumer, not API owner — polls the agent's existing scoreboard endpoints. No new API surface required on the agent.
- Graceful degradation — if an endpoint is unavailable (agent doesn't have maturation enabled, work items disabled, etc.), that panel shows "N/A" instead of erroring.
| Endpoint | Source | Data |
|---|---|---|
GET /v1/agent/trust |
fipsagents.server.trust_routes |
Trust level, score, history |
GET /v1/agent/maturation |
fipsagents.server.trust_routes |
Stage, permissions, promotion progress |
GET /v1/agent/skills |
fipsagents.server.trust_routes |
Bundled + learned skills |
GET /v1/agent/capabilities |
fipsagents.server.trust_routes |
Discovered capabilities |
GET /v1/work-items/stats |
fipsagents.server.trust_routes |
Queue depth by status |
GET /healthz |
fipsagents.server.app |
Agent health |
Apache-2.0