Default route for background jobs and prompt utilities.
AI Configuration
One place for coding agents, gateways, AI providers, model health, testing, and Wall-E routing.
Used for Wall-E chat, reflection, and scheduled jobs.
Runtime is a coding agent; model route comes from AI Providers.
Coding Agents
Session runners are configured separately from provider/model access.
Runs Claude coding sessions.
Runs OpenAI coding sessions.
Runs Google/Gemini sessions.
Uses CTM/Wall-E model routes.
Gateways
Shared routing layers for model discovery and provider access.
Portkey Gateway
Connected · 33 Portkey providers · 2,483 imported models
AI Providers
Keep the existing model/provider information, tests, rescans, credentials, and route policy here.
42 models · 2 routes · Used by Wall-E jobs
1 direct · 1 Portkey · provider-level policy
Credential set · provider API · 42 models
Credential set · gateway route · 38 models
175 models · 3 routes · CTM default
2 direct · 1 Portkey · provider-level policy
Credential set · provider API · 108 models
27 models · Portkey-only
Created from Portkey catalog discovery.
Gateway route · 27 models
Model Catalog
Long model inventory stays available for inspection, search, pricing, route source, and ambiguity debugging.
Settings Mock
Settings keeps a summary/deep link, not a second AI provider editor.
AI summary
Open the AI Configuration page for provider cards, models, testing, gateways, and defaults.