Platform
System primitives: AI inference connectors, federated AEGIS hubs, Cortex/Erin/Lexicon stack, container topology.
AI inference connectors
Llm-connector instances proxying ollama / cloud providers. Configured in hub-api env.
No connectors registered.
AEGIS federation peers
Other hubs in the federation. Each hub keeps its own credentials, RLS, and TURN.
No federated peers (or registry empty).
Cortex / Erin / Lexicon
The AGI runtime, her workplace, and her context server.
- Erinerin.tsho.meAGI runtime
- Cortexcortex.tsho.uscontext server
- Lexicon (hub-native)/lexicon/*workplace
Hub identity & auth
- Hub URL
- hub.jtnt.us
- Auth realm
- JTNT @ auth.tsho.me
- Database role
- jtnt_hub_app (NOSUPERUSER)
- RLS
- enforcing
- Vault crypto
- aes-256-gcm
Restart, scale, and config changes still require shell access on hub.jtnt.us:
docker compose restart hub-api.