이대희
c548c5d49f
Fixes Claude API thinking block requirement
...
Addresses a Claude API requirement where assistant messages with tool use must have a thinking block when thinking is enabled.
This commit injects an empty thinking block into assistant messages that include tool use but lack a thinking block. This ensures compatibility with the Claude API when the thinking feature is enabled.
2026-02-02 14:04:29 +09:00
hkfires
c8c27325dc
feat(thinking): enable thinking toggle for qwen3 and deepseek models
...
Fix #1245
2026-01-28 09:54:05 +08:00
hkfires
239a28793c
feat(claude): clamp thinking budget to max_tokens constraints
2026-01-19 16:32:20 +08:00
hkfires
c421d653e7
refactor(claude): move max_tokens constraint enforcement to Apply method
2026-01-19 15:50:35 +08:00
hkfires
4ad6189487
refactor(thinking): extract antigravity logic into a dedicated provider
2026-01-15 19:08:22 +08:00
hkfires
33d66959e9
test(thinking): remove legacy unit and integration tests
2026-01-15 13:06:40 +08:00
hkfires
7f1b2b3f6e
fix(thinking): improve model lookup and validation
2026-01-15 13:06:40 +08:00
hkfires
40ee065eff
fix(thinking): use static lookup to avoid alias issues
2026-01-15 13:06:40 +08:00
hkfires
72f2125668
fix(executor): properly handle thinking application errors
2026-01-15 13:06:39 +08:00
hkfires
e8f5888d8e
fix(thinking): fix auth matching for thinking suffix and json field conflicts
2026-01-15 13:06:39 +08:00
hkfires
0b06d637e7
refactor: improve thinking logic
2026-01-15 13:06:39 +08:00
