Back to dashboard | Back to daily
AI Runtime Status
AI evidence page
Return to /daily for daily decisions
- AI runs collection, analysis, drafting, and queueing from the daily pipeline.
- Humans only approve, reject, or request edits from Today's decisions.
- AI runtime status explains model/provider execution evidence. Daily operators do not need this unless /daily shows an AI issue.
Use this page to understand what AI did; do not use it as the normal daily operating path.
AI runtime evidence page
Return to /daily for daily decisions
Use this page to inspect AI provider mode, probe output, and fallback status only. AI diagnostics and logs are evidence; humans decide from the daily console.
Verifies whether GrowthOS is using a live AI provider or deterministic fallback.
Mode
provider_configured
Providers
1
Default
deepseek
Last Probe
OK
AI provider runtime is verified
Continue daily operations; rerun the probe after rotating provider keys.
Last probe: 2026-07-19T13:07:28.900Z
Provider: deepseek / deepseek-chat
Fallback: no
GrowthOS AI runtime is active and ready to support ecommerce operations.
Providers
| Provider | Model | Status | Action |
|---|---|---|---|
| openai | gpt-4.1-mini | missing key | Set OPENAI_API_KEY and restart GrowthOS. |
| anthropic | claude-3-5-sonnet-latest | missing key | Set ANTHROPIC_API_KEY and restart GrowthOS. |
| gemini | gemini-1.5-pro | missing key | Set GEMINI_API_KEY and restart GrowthOS. |
| deepseek | deepseek-chat | configured | Ready for AI-assisted jobs. |
| qwen | qwen-plus | missing key | Set QWEN_API_KEY and restart GrowthOS. |
Safeguards
| The read status endpoint never calls an AI provider. |
| AI runtime probe is an admin diagnostic and requires the GrowthOS admin token. |
| The probe sends a short generic prompt and no store secrets. |
| Fallback mode remains operational when no provider key is configured. |
Links
AI runtime probes are admin diagnostics and require the GrowthOS admin token; daily operators should continue from /daily.