Commit Graph

388 Commits

Author SHA1 Message Date
jxxghp
314ac3903c fix(agent): 识别 DeepSeek image_url 反序列化错误,修复图片不支持提示 (#5863) 2026-05-30 20:15:24 +08:00
jxxghp
492e3c333b fix: unify User-Agent header usage with settings.USER_AGENT 2026-05-30 09:49:29 +08:00
jxxghp
762a7fbba7 fix: simplify search web providers 2026-05-28 20:11:31 +08:00
jxxghp
10290ca17b 优化 agent 网络搜索工具 2026-05-28 17:22:52 +08:00
jxxghp
cc3e062262 固定 Agent 工具消息类型 2026-05-28 13:34:53 +08:00
jxxghp
bf4f5f8744 优化 Agent 核心提示词结构 2026-05-28 11:22:56 +08:00
jxxghp
f8f06a602a 加强 Agent 核心提示词安全约束 2026-05-28 11:10:08 +08:00
jxxghp
0e5c592862 refactor: centralize MoviePilot server helper 2026-05-27 12:56:45 +08:00
jxxghp
5b6b4c9744 将发送消息工具设为必选工具 2026-05-27 08:48:56 +08:00
jxxghp
990a28b51b 调整 Agent 心跳任务报告方式 2026-05-27 08:41:40 +08:00
jxxghp
1f7fb304dd feat: add LLM proxy toggle 2026-05-27 06:57:09 +08:00
jxxghp
896631d63e fix(agent): enable voice replies for supported channels 2026-05-26 20:14:56 +08:00
jxxghp
31554bdcb5 fix agent tool call history ordering 2026-05-26 16:42:52 +08:00
jxxghp
e00e18f31e fix: transcode unsupported stt audio formats 2026-05-26 10:00:32 +08:00
jxxghp
c7965edd47 fix: use responses api for chatgpt reasoning models 2026-05-26 09:31:00 +08:00
jxxghp
821bd3decd fix: use explicit LLM user agent config 2026-05-26 08:24:58 +08:00
jxxghp
b65c8dcfe0 feat: support llm user agent 2026-05-26 08:20:02 +08:00
jxxghp
d940373f6b 将所有agent工具的explanation字段改为可选
修复Pydantic验证错误:QueryTransferHistoryInput的explanation字段为必需,但用户未提供。
修改了74个工具文件,将explanation字段从必需改为可选,默认值为None。
2026-05-25 16:40:43 +08:00
jxxghp
0e8bcb4df6 fix: patch gemini thought_signature enforcement to cover all function calls
The upstream _parse_chat_history enforcement code uses a first_fc_seen
flag that only adds DUMMY_THOUGHT_SIGNATURE to the first function_call
without thought_signature. Parallel function calls (position 2+) remain
unpatched, causing Gemini API 400 errors for all Gemini 2.5+ models.

Additionally, _is_gemini_3_or_later only matches 'gemini-3', missing
Gemini 2.5 models entirely.

This patch:
1. Extends _is_gemini_3_or_later to also match gemini-2.5 models
2. Wraps _parse_chat_history to ensure ALL function_call parts in ALL
   model messages have thought_signature (not just the first one)
2026-05-25 13:53:02 +08:00
jxxghp
766d2699ea feat: support MiniMax audio provider 2026-05-25 11:42:57 +08:00
jxxghp
477c49587c feat(agent): log tool execution result summary and truncate if too long 2026-05-25 08:50:44 +08:00
jxxghp
5532f14efb feat(agent): inject plugin installation directory into system prompt 2026-05-25 08:21:19 +08:00
jxxghp
b08c335bb4 fix: 优化 execute_command 工具的 rm -rf 拦截逻辑,仅禁止删除根目录或一级目录 2026-05-25 08:00:35 +08:00
jxxghp
c7670e5cc8 更新 execute_command.py 2026-05-25 07:51:30 +08:00
jxxghp
a725789045 feat: add agent token provider events 2026-05-25 07:32:36 +08:00
jxxghp
ac3432c54f feat: support TMDB episode group (g=) in explicit media tags and custom identifiers
- Add episode_group (g=) parameter parsing to explicit media tags in both Python and Rust metainfo parsers
- Propagate episode_group through MetaInfo, MetaBase, MediaInfo, and context models
- Update SKILL.md and update_custom_identifiers.py docs to describe episode group usage
- Add tests for episode_group recognition in metainfo and chain recognition logic
2026-05-24 23:32:27 +08:00
jxxghp
a97e3ea092 feat: improve agent execution error handling with user-friendly messages and sensitive info sanitization 2026-05-24 22:14:10 +08:00
jxxghp
79539760da fix: bound long-lived cache state 2026-05-24 18:03:42 +08:00
jxxghp
dc73d61682 feat: add china operator llm providers 2026-05-24 09:30:41 +08:00
jxxghp
134c441754 fix: replace fast jieba dependency 2026-05-23 12:59:33 +08:00
jxxghp
00fc8b2f53 chore: update text processing dependencies 2026-05-23 12:18:59 +08:00
jxxghp
b26bf2a019 fix: stop agent after user choice prompt 2026-05-23 00:15:35 +08:00
jxxghp
a74f04a149 fix: simplify message typing lifecycle 2026-05-23 00:11:56 +08:00
jxxghp
2b5528c0ac fix: keep agent typing status while queued 2026-05-22 16:46:25 +08:00
jxxghp
737bcb5c62 refactor(agent): move feedback issue flow into skill scripts 2026-05-21 19:22:27 +08:00
InfinityPacer
0245c8db80 feedback-issue: 拆三步、入口意图门、消息可靠性、日志脱敏与噪音过滤 (#5810) 2026-05-21 13:57:12 +08:00
jxxghp
ee9eced2f1 fix: avoid blocking event loop during plugin install 2026-05-21 09:16:42 +08:00
InfinityPacer
ad7cce72f4 新增 feedback-issue Agent skill:把用户反馈整理为上游 Issue (#5799) 2026-05-20 20:10:03 +08:00
jxxghp
c52ccaf75f feat: add plugin system version compatibility checks 2026-05-20 19:55:44 +08:00
jxxghp
c661bc4764 perf: reduce agent shell command probing 2026-05-20 18:50:59 +08:00
jxxghp
5fe5523d13 fix: prefer rg in agent prompts 2026-05-20 15:31:02 +08:00
jxxghp
5d02550874 fix: cache available shell command discovery 2026-05-19 18:45:52 +08:00
jxxghp
c4eb4d9b95 feat: inject available shell commands into agent prompt 2026-05-19 18:16:09 +08:00
Album
9e9c398177 feat: 新增集数定位模板生成接口 (#5785) 2026-05-19 07:18:28 +08:00
jxxghp
9076acc52e feat: add managed agent command sessions 2026-05-18 20:17:59 +08:00
jxxghp
3ee601574c fix: reduce low-risk pylint issues 2026-05-17 08:01:39 +08:00
jxxghp
0ee9fec1d2 feat(browser): migrate to CloakBrowser for browser emulation and streamline dependency management
- Replace Playwright-based browser emulation with CloakBrowser as default
- Update config to support CloakBrowser options and humanization presets
- Refactor browser helper to use CloakBrowser context and remove cf_clearance dependency
- Update Dockerfile, entrypoint, and update scripts to install CloakBrowser runtime
- Ensure CloakBrowser kernel is pre-installed during local setup and dependency updates
- Add tests for CloakBrowser integration and legacy compatibility
2026-05-16 20:51:38 +08:00
jxxghp
b2a18f9ae4 feat(message-processing-status): unified processing status indicator for Telegram, Slack, Discord, Feishu
- Add ChannelCapability.PROCESSING_STATUS and capability detection for supported channels
- Implement mark_message_processing_started/finished in Telegram, Slack, Discord, Feishu modules
  - Telegram: manage typing lifecycle with max duration and explicit stop
  - Slack: add/remove reaction as processing indicator
  - Discord: start/stop typing indicator with async task management
  - Feishu: add/remove reaction for processing status
- Refactor message chain to invoke processing status hooks for supported channels
- Ensure processing status is properly finished on sync and async message handling paths
- Add tests for processing status lifecycle and capability detection across channels
2026-05-15 12:45:41 +08:00
jxxghp
23784f614b feat: add user-friendly handling for unsupported image input errors in agent execution 2026-05-14 20:36:14 +08:00
jxxghp
7b27b7fd16 feat: add extensible agent audio capabilities 2026-05-14 19:37:13 +08:00