LLM INFERENCE GATEWAY // OPENAI-COMPATIBLE
INTELLIGENCE PROVIDER CONFIGURATION
STATUS:ACTIVE ROUTING
Default gateway endpoint for autonomous triage, 3-bullet briefs, and ambiguity scanning.
Model tag dispatched in the inference request payload (e.g., indunix-1-core).
Stored securely on the backend server; never exposed to public client requests.
Inference Temperature0.20
Low temperature (0.1–0.3) ensures deterministic, reproducible acceptance criteria scoring.
Tokens allocation cap per brief generation (default: 1024).
GATEWAY CONNECTIVITY AUDIT
Executes an authenticated probe request to verify bearer credentials and measure edge roundtrip latency.