b163f8ed9e
Fixed : #1004
...
feat(translator): add function name to response output item serialization
- Included `item.name` in the serialized response output to enhance output item handling.
2026-01-15 03:27:00 +08:00
83e5f60b8b
fix(kiro): scale description compression by needed size
...
Compute a size-reduction based keep ratio and use it to trim
tool descriptions, avoiding forced minimum truncation when the
target size already fits. This aligns compression with actual
payload reduction needs and prevents over-compression.
2026-01-14 16:22:46 +08:00
5b433f962f
feat(kiro): 实现动态工具压缩功能
...
## 背景
当 Claude Code 发送过多工具信息时，可能超出 Kiro API 请求限制导致 500 错误。
现有的工具描述截断（KiroMaxToolDescLen = 10237）只能限制单个工具的描述长度，
无法解决整体工具列表过大的问题。
## 解决方案
实现动态工具压缩功能，采用两步压缩策略：
1. 先检查原始大小，超过 20KB 才进行压缩
2. 第一步：简化 input_schema，只保留 type/enum/required 字段
3. 第二步：按比例缩短 description（最短 50 字符）
4. 保留全部工具和 skills 可调用，不丢弃任何工具
## 新增文件
- internal/translator/kiro/claude/tool_compression.go
- calculateToolsSize(): 计算工具列表的 JSON 序列化大小
- simplifyInputSchema(): 简化 input_schema，递归处理嵌套 properties
- compressToolDescription(): 按比例压缩描述，支持 UTF-8 安全截断
- compressToolsIfNeeded(): 主压缩函数，实现两步压缩策略
- internal/translator/kiro/claude/tool_compression_test.go
- 完整的单元测试覆盖所有新增函数
- 测试 UTF-8 安全性
- 测试压缩效果
## 修改文件
- internal/translator/kiro/common/constants.go
- 新增 ToolCompressionTargetSize = 20KB (压缩目标大小阈值)
- 新增 MinToolDescriptionLength = 50 (描述最短长度)
- internal/translator/kiro/claude/kiro_claude_request.go
- 在 convertClaudeToolsToKiro() 函数末尾调用 compressToolsIfNeeded()
## 测试结果
- 70KB 工具压缩至 17KB (74.7% 压缩率)
- 所有单元测试通过
## 预期效果
- 80KB+ tools 压缩至 ~15KB
- 不影响工具调用功能
2026-01-14 11:07:07 +08:00
5977af96a0
fix(antigravity): prevent corrupted thought signature when switching models
...
When switching from Claude models (e.g., Opus 4.5) to Gemini models
(e.g., Flash) mid-conversation via Antigravity OAuth, the client-provided
thinking signatures from Claude would cause "Corrupted thought signature"
errors since they are incompatible with Gemini API.
Changes:
- Remove fallback to client-provided signatures in thinking block handling
- Only use cached signatures (from same-session Gemini responses)
- Skip thinking blocks without valid cached signatures
- tool_use blocks continue to use skip_thought_signature_validator when
no valid signature is available
This ensures cross-model switching works correctly while preserving
signature validation for same-model conversations.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com >
2026-01-13 18:24:05 +08:00
216dafe44b
Merge branch 'router-for-me:main' into main
2026-01-12 00:27:15 +08:00
220ca45f74
fix(codex): only override instructions when upstream provides them
2026-01-11 15:52:21 +08:00
70a82d80ac
fix(codex): only override instructions in responses for OpenCode UA
2026-01-11 15:19:37 +08:00
ac626111ac
feat(codex): add OpenCode instructions based on user agent
2026-01-11 13:36:35 +08:00
5bb9c2a2bd
Add candidate count parameter to OpenAI request
2026-01-10 18:50:13 +08:00
0b5bbe9234
Add candidate count handling in OpenAI request
2026-01-10 18:49:29 +08:00
14c74e5e84
Handle 'n' parameter for candidate count in requests
...
Added handling for the 'n' parameter to set candidate count in generationConfig.
2026-01-10 18:48:33 +08:00
6448d0ee7c
Add candidate count handling in OpenAI request
2026-01-10 18:47:41 +08:00
b0c17af2cf
Enhance Gemini to OpenAI response conversion
...
Refactor response handling to support multiple candidates and improve parameter management.
2026-01-10 18:46:25 +08:00
514b9bf9fc
Merge origin/main into pr-92
2026-01-10 01:12:22 +08:00
4d7f389b69
Fixed : #941
...
fix(translator): ensure fallback to valid originalRequestRawJSON in response handling
2026-01-10 01:01:09 +08:00
1906ebcfce
Merge branch 'main' into plus
2026-01-09 21:52:24 +08:00
ee62ef4745
refactor(logging): clean up oauth logs and debugs
2026-01-09 11:20:55 +08:00
d47b7dc79a
refactor(response): enhance parameter handling for Codex to Claude conversion
2026-01-09 05:20:19 +08:00
3d01b3cfe8
Merge pull request #553 from XInTheDark/fix/builtin-tools-web-search
...
fix(translator): preserve built-in tools (web_search) to Responses API
2026-01-09 04:40:13 +08:00
4794645dec
Merge branch 'main' into plus
2026-01-06 23:20:39 +08:00
8f8dfd081b
Merge pull request #850 from can1357/main
...
feat(translator): add developer role support for Gemini translators
2026-01-05 11:27:24 +08:00
478aff1189
Merge branch 'router-for-me:main' into main
2026-01-05 09:26:23 +08:00
8edbda57cf
feat(translator): add thoughtSignature to node parts for Gemini and Antigravity requests
...
Enhanced node structure by including `thoughtSignature` for inline data parts in Gemini OpenAI, Gemini CLI, and Antigravity request handlers to improve traceability of thought processes.
2026-01-05 09:25:17 +08:00
6762e081f3
feat(translator): add developer role support for Gemini translators
...
Treat OpenAI's "developer" role the same as "system" role in request
translation for gemini, gemini-cli, and antigravity backends.
2026-01-03 21:01:01 +01:00
771fec9447
Merge branch 'main' into plus
2026-01-04 01:38:40 +08:00
7815ee338d
fix(translator): adjust message_delta emission boundary in Claude-to-OpenAI conversion
...
Fixed incorrect boundary logic for `message_delta` emission, ensuring proper handling of usage updates and `emitMessageStopIfNeeded` within the response loop.
2026-01-04 01:36:51 +08:00
ba486ca6b7
Merge branch 'router-for-me:main' into main
2026-01-01 21:03:16 +08:00
2d2f4572a7
fix(translator): remove unnecessary whitespace trimming in reasoning text collection
2026-01-01 12:39:09 +08:00
8f4c46f38d
fix(translator): emit tool_result messages before user content in Claude-to-OpenAI conversion
2026-01-01 11:11:43 +08:00
b6ba51bc2a
feat(translator): add thinking block and tool result handling for Claude-to-OpenAI conversion
2026-01-01 09:41:25 +08:00
e85c9d9322
Merge branch 'main' into plus
2025-12-30 23:33:35 +08:00
c6b0e85b54
Fixed : #790
...
fix(gemini): include full text in response output events
2025-12-30 20:44:13 +08:00
cf369d4684
Merge branch 'router-for-me:main' into main
2025-12-29 22:41:44 +08:00
d8fc485513
fix(translators): correct key path for system_instruction.parts in Claude request logic
2025-12-29 11:54:26 +08:00
d473c952fb
Merge branch 'main' into plus
2025-12-28 00:56:04 +08:00
7646a2b877
Fixed : #749
...
fix(translators): ensure `gjson.String` content is non-empty before setting `parts` in OpenAI request logic
2025-12-28 00:54:26 +08:00
d35152bbef
Merge branch 'router-for-me:main' into main
2025-12-27 22:03:50 +08:00
c281f4cbaf
Fixed : #747
...
fix(translators): rename and integrate `usageMetadata` as `cpaUsageMetadata` in Claude processing logic
2025-12-27 22:02:11 +08:00
60936b5185
Merge branch 'router-for-me:main' into main
2025-12-27 03:57:03 +08:00
72274099aa
Fixed : #738
...
fix(translators): refine prompt token calculation by incorporating cached tokens in Claude response handling
2025-12-27 03:56:11 +08:00
b7f7b3a1d8
Merge branch 'router-for-me:main' into main
2025-12-27 01:26:33 +08:00
dcae098e23
Fixed : #736
...
fix(translators): handle gjson string types in Claude request processing to ensure consistent content parsing
2025-12-27 01:25:47 +08:00
9fe6a215e6
Merge branch 'router-for-me:main' into main
2025-12-26 05:10:19 +08:00
33e53a2a56
fix(translators): ensure correct handling and output of multimodal assistant content across request handlers
2025-12-26 05:08:04 +08:00
3f949b7f84
Merge pull request #704 from tinyc0der/add-index
...
fix(openai): add index field to image response for LiteLLM compatibility
2025-12-25 21:35:12 +08:00
a7fc2ee4cf
refactor(image): avoid using json.Marshal
2025-12-25 14:21:01 +07:00
ed57d82bc1
Merge branch 'router-for-me:main' into main
2025-12-24 23:31:09 +08:00
06ad527e8c
Fixed : #696
...
fix(translators): adjust prompt token calculation by subtracting cached tokens across Gemini, OpenAI, and Claude handlers
2025-12-24 23:29:18 +08:00
671558a822
fix(openai): add index field to image response for LiteLLM compatibility
...
LiteLLM's Pydantic model requires an index field in each image object.
Without it, responses fail validation with "images.0.index Field required".
2025-12-24 17:43:31 +07:00
c169b32570
refactor(kiro): Remove unused variables in OpenAI translator
...
Remove dead code that was never used:
- toolCallIDToName map: built but never read from
- seenToolCallIDs: declared but never populated, only suppressed with _
2025-12-24 15:10:35 +07:00
Luis Pater