upgrade: pin LiteLLM v1.108.0 (candidate for 2026-11-08 window)

- Dockerfile FROM v1.103.2 -> v1.108.0; manifest upstreamVersion follows.
- patch_expose_cost_headers.py: upstream 1.108.0 refactored the CORS
  middleware call (expose_headers now also spreads
  trace_response_expose_headers()), so the 1.103.x anchor no longer
  matched and the patch would have SILENTLY stopped applying. The script
  now handles both anchor shapes (1.103.x first, 1.108.x fallback).
- All 8 build patches verified against the real v1.108.0 sources: every
  anchor matched, every patched file compiles, re-runs are idempotent.
- 36 new migrations, all additive (no unguarded DROP/DELETE/UPDATE);
  budget_reset_at unchanged on LiteLLM_TeamTable; SpendLogs gains only
  billing_agent_id.
This commit is contained in:
inference-bot committed 2026-11-06 03:55:36 +00:00
1 parent e161db9ec8
commit 9e25578ee6
4 files changed
+38 -11

No files matched your search

+1 -1
View File
@@ -5,7 +5,7 @@
"description": "LiteLLM AI Gateway \u2014 OpenAI-compatible proxy with per-user API keys, usage tracking, rate limiting, and model routing. Packaged for Cloudron with SSO/OIDC and PostgreSQL.",
"tagline": "AI gateway for cooperative inference",
"version": "1.0.0",
"upstreamVersion": "1.103.2",
"upstreamVersion": "1.108.0",
"healthCheckPath": "/health/readiness",
"httpPort": 4000,
"memoryLimit": 2147483648,