✔ actual config HTTP failure preserves the committed selection, config and volatile credential before a successful retry (1275.6896ms) ✔ model registry keeps original credentials/endpoint intact, rejects stale edits, and forgets only volatile keys on restart (1061.908ms) ✔ config import previews exact revisions, keeps secrets volatile, syncs only when enabled and preserves manual edits on conflicts (902.1122ms) ✔ real HTTP pool fails over 429 to another model and never repeats completed tools; attempts seal actual request bytes and Run identity (3585.9279ms) ✔ pool rotation and weights select members on new Runs; disabled and missing credentials are excluded without requests (5134.0748ms) ✔ 401 blocks the credential revision; updating the key restores availability without leaking error echoes (2257.5361ms) ✔ queued request cancellation and in-flight disable release only owned leases and do not dispatch another HTTP request (996.2576ms) ✔ compiled model-center HTTP API persists public profiles, blocks forged origins, and local operations never invoke a model (2226.8208ms) ✔ model registry disk failures restore routing, pool profiles and credential references without partial updates (415.037ms) ✔ model selection and saved configuration commit together and failures preserve the original mode and endpoint (572.4144ms) ✔ configuration imports roll back the entire batch and retain one usable preview after validation or commit failure (597.6559ms) ✔ identity distinguishes missing, invalid, mixed, exact and explicitly declared aliases without guessing (1.1285ms) ✔ chat-completions JSON rejects a different model before tool dispatch, retaining source and reported usage (2042.9059ms) ✔ chat-completions SSE rejects a different model before tool dispatch, retaining source and reported usage (1815.0182ms) ✔ responses JSON rejects a different model before tool dispatch, retaining source and reported usage (1893.7824ms) ✔ responses SSE rejects a different model before tool dispatch, retaining source and reported usage (1886.6307ms) ✔ record policy accepts missing identity but keeps it explicit, and strict declared mapping is frozen in the consumption (1.11ms) ✔ mixed streamed identities cannot pass strict mode, and interrupted streams retain observed identifiers without authenticating them (1.1194ms) ✔ identity refusal never fails over to another pool member and raw response model cannot leak known credentials (1918.3608ms) ✔ strict refusal remains visible even when the same original response exhausts the budget (2028.3125ms) ✔ strict identity crosses the controlled engine bridge with its original source and zero tool dispatch (12290.6912ms) ✔ frozen direct connection policy and delivery reports preserve response identity without filling historical gaps (2398.9469ms) ✔ capability routing skips text-only members before acquiring a lease and records exact refusal in the original Run (1231.814ms) ✔ all incompatible members reject with no network request, no lease, preserved Task and explicit non-dispatched ledger (823.0826ms) ✔ explicit synthetic tool probe verifies only actual matching call, never executes it and loses verification on revision changes (623.0905ms) ✔ plain text probe cannot verify tools; wrong tool nonce and client supplied evidence cannot claim verified support (641.2864ms) ✔ a disabled connection invalidates an in-flight probe even when toggle mutates its original registry object (711.4759ms) ✔ selected connection owns the actual endpoint, credential, wire model and mismatch even when legacy is configured for GPT (2410.5816ms) ✔ budget-only API and demo/real selection preserve routes, reject injected endpoint fields and publish authoritative selection without keys (2646.3163ms) ✔ another session cannot change default model, mode or budgets while a request is active (3170.6177ms) ℹ tests 30 ℹ suites 0 ℹ pass 30 ℹ fail 0 ℹ cancelled 0 ℹ skipped 0 ℹ todo 0 ℹ duration_ms 30630.6759