zhuchaowe/qwen_agent
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
631 Commits 4 Branches 0 Tags 68 MiB
2964326cb8
Commit Graph

11 Commits

Author SHA1 Message Date
朱潮
d1e31953b0 merge guideline-assistant-message-error 2026-04-09 15:05:11 +08:00
朱潮
05e22391ed merge FACT_RETRIEVAL_PROMPT.md 2026-04-03 10:34:09 +08:00
朱潮
845815866e Merge branch 'feature/pre-memory-prompt' into dev 2026-04-02 16:49:32 +08:00
朱潮
7da8466b3d ✨ feat: 记忆排除规则新增设备状态查询、故障报告、联系方式查找 
- 新增排除类别：设备/设施状态查询与结果、Bug/故障报告、联系方式查找
- 新增3个日语负例 few-shot examples

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-02 16:48:11 +08:00
朱潮
76e742e7b3 merge FACT_RETRIEVAL_PROMPT.md 2026-04-02 16:45:29 +08:00
朱潮
bb23d230d1 ✨ feat: 优化记忆提示词，过滤操作性/查询性动作 
- 新增 EXCLUDE 排除规则，过滤查询动作、设备操作、一次性指令等不应记忆的信息
- 新增7个日语负例 few-shot examples（查询员工信息、操作灯光、设置空调等）
- 新增核心判断规则：只记忆用户是谁，不记忆用户让助手做了什么
- 同步合并 dev 分支的 plain language 规则和语义完整性改进

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-02 16:44:04 +08:00
朱潮
738a904a6f merge 2026-04-02 11:25:20 +08:00
autobee-sparticle
85519da5a5
fix(memory): 改进 Memory 提取 prompt 使用大白话 (#20) 
将 fact extraction prompt 中的技术性格式改为自然语言：
- "Contact: [name] (relationship, referred as [nick])" → "[name] is a [relationship], also called [nick]"
- 移除 "DEFAULT when user says" 等技术标记
- 添加 Plain Language Rule 明确要求输出通俗易懂的文本
- 更新所有示例为自然语言风格

Co-authored-by: zhuchao <zhuchaowe@163.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-26 10:14:13 +09:00
autobee-sparticle
deb78a7625
fix: improve memory extraction for colloquial/informal speech (#16) 
* chore: add .worktrees/ to .gitignore

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat(CI): 添加 onprem-dev 环境的构建和部署配置

在 CircleCI 配置中新增 onprem-dev 环境的 build-and-push 和 deploy 任务，部署到 cluster-for-B 的 onprem-dev 命名空间

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* fix: improve memory extraction for colloquial/informal speech

Add semantic completeness rules and multilingual few-shot examples
to FACT_RETRIEVAL_PROMPT to prevent truncated or semantically incorrect
memory extraction. Specifically addresses Japanese casual speech where
particles (が, を, に) are often omitted.

Closes sparticleinc/mygpt-frontend#2125

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: zhuchao <zhuchaowe@163.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: shuirong <shuirong1997@icloud.com>
 2026-03-17 10:37:49 +09:00
朱潮
17f4217f29 添加了同名人员处理说明 2026-03-02 16:06:58 +08:00
朱潮
60bd1e4042 ✨ feat(prompt): add fact retrieval prompt with contact tracking 
Add comprehensive prompt for extracting and storing user facts from
conversations, with special focus on relationship/contact tracking:
- Full name and nickname association (e.g., "Mike" → "Michael Johnson")
- Relationship context recording (family, friend, colleague, etc.)
- Multi-language name support
- Few-shot examples for various fact extraction scenarios

Generated with [Claude Code](https://claude.ai/code)
via [Happy](https://happy.engineering)

Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Happy <yesreply@happy.engineering>
 2026-01-21 13:22:17 +08:00