Mario Zechner 8df22faedf fix(ai): ensure maxTokens > thinkingBudget for Claude thinking models ... Claude requires max_tokens > thinking.budget_tokens. When caller specifies a small maxTokens (e.g. compaction with ~13k tokens) and reasoning is enabled with high budget (16k tokens), the constraint was violated. Fix: In mapOptionsForApi, add thinkingBudget on top of caller's maxTokens (capped at model.maxTokens). If still not enough room, reduce thinkingBudget to leave space for output. Applied to both anthropic-messages and google-gemini-cli APIs. Also adds test utilities for OAuth credential resolution and tests for compaction with thinking models. fixes #413		2026-01-03 02:45:30 +01:00
..
providers	Fix google-vertex build	2026-01-03 01:11:03 +01:00
utils	Use dynamic import for http module in OAuth files	2025-12-30 22:42:21 +01:00
cli.ts	WIP: Add CLI for OAuth login, update README	2025-12-25 01:09:27 +01:00
index.ts	Add Vertex AI provider with ADC support	2026-01-03 01:11:03 +01:00
models.generated.ts	Fix google-vertex build	2026-01-03 01:11:03 +01:00
models.ts	fix: use consistent model comparison including provider	2025-12-25 22:10:08 +01:00
stream.ts	fix(ai): ensure maxTokens > thinkingBudget for Claude thinking models	2026-01-03 02:45:30 +01:00
types.ts	Add Vertex AI provider with ADC support	2026-01-03 01:11:03 +01:00