AegonDivel Waguri online
Unified control plane for Hermes on aegondivel.id. Switch the global model in ~/.hermes/config.yaml with one click, proxy chat through ai.aegondivel.id/v1 (9router at 127.0.0.1:20128), and monitor system health in real time.
Global Model Switch — POST /api/hermes/model
Playground — POST /api/chat
Health — GET /api/health
Models — GET /api/models 410
CPU Load (1m)
--
--
Memory
--
--
Disk /
--
--
9router
--
offline or checking
Global Model Switch — 410 models
Current default
loading
provider: custom
Grouped by prefix — click switch to update ~/.hermes/config.yaml → model.default globally. Each switch calls POST /api/hermes/model.
Loading models from GET /api/models...
Playground Chat
model: --System — Playground proxies to ai.aegondivel.id/v1 via POST /api/chat. Pick a model, then send a message. Response streams from 9router.
Uses current default model if none selected. Payload: { model, messages } forwarded to 127.0.0.1:20128/v1/chat/completions with bearer from ~/.hermes/.env. Set max_tokens/temperature in code if needed.
Quick curl snippet
curl -s http://127.0.0.1:3001/api/hermes/model \
-H "Content-Type: application/json" \
-d '{"model":"skd/meta/muse-spark-1.2-contributor"}'
# list models
curl -s http://127.0.0.1:3001/api/models | head -c 600
# chat proxy
curl -s http://127.0.0.1:3001/api/chat \
-H "Content-Type: application/json" \
-d '{"model":"skd/meta/muse-spark-1.2-contributor","messages":[{"role":"user","content":"hello"}]}'
Snippets target the Dashboard backend on :3001. The switch endpoint writes ~/.hermes/config.yaml on the VPS immediately.
How it works
aegondivel.id · ai.aegondivel.id → 127.0.0.1:201281. Browse and switch
Search the 410-model catalog grouped by prefix (skd, skd2, cx, tabi, vantis, f, ar, cup, th, qd, go, meta). Hit switch to run POST /api/hermes/model which updates model.default in ~/.hermes/config.yaml.
Search the 410-model catalog grouped by prefix (skd, skd2, cx, tabi, vantis, f, ar, cup, th, qd, go, meta). Hit switch to run POST /api/hermes/model which updates model.default in ~/.hermes/config.yaml.
2. Playground verification
Chat via POST /api/chat to confirm the new default works end-to-end through 9router. The proxy uses the bearer from ~/.hermes/.env.
Chat via POST /api/chat to confirm the new default works end-to-end through 9router. The proxy uses the bearer from ~/.hermes/.env.
3. Health and ops
Health cards poll GET /api/health for CPU, RAM, disk, and live 9router model count. Curl snippets are one-click copyable for ops and CI.
Health cards poll GET /api/health for CPU, RAM, disk, and live 9router model count. Curl snippets are one-click copyable for ops and CI.