feat: 批量采集管道(Temu→1688 选品比价)与桌面级 computer_use#8
Merged
Conversation
围绕"把 Temu 已发布商品批量采集到 WPS 成本核算表"的完整能力,含支撑设施与 Windows/跨平台健壮性修复。 新增能力: - app/collect/:确定性两段式采集管道。pipeline.py 把第二段 agent 自由循环 换成逐商品确定性例程 + 2 次单发大模型(视觉挑同款 + 文本读价),单商品 token 从十万级压到两次小输入;image_extract.py 用生图模型把营销拼图重绘成 白底单品图仅作图搜查询图(判同款仍用原图,误差只影响 recall 不影响精度)。 - app/tool/wps_excel_tool.py:WPS DISPIMG 表安全写入。全程 zip/XML 增量改, 绝不用 openpyxl(其 save 会删 cellimages.xml 毁掉所有嵌入商品图)。 - app/tool/computer_use_tool.py:系统桌面级 computer use(整屏截图 + 鼠标/键盘/ 窗口控制),与 browser_use 互补处理浏览器够不到的原生对话框/桌面应用。 - batch_collect.py:两段式批量驱动。第一段命中 Temu 列表接口 0-LLM 枚举全量 SPU;第二段对每个未入库 SPU 清空记忆独立跑,幂等可断点续跑。 - browser_use_tool.py:paste_image(剪贴板 + 真实 Ctrl+V 以图搜图,绕开会冻死 自动化的原生文件对话框)、close_tabs、execute_js 批量读字段等。 - 提示词红线:WPS 表操作、1688 采购价/重量采集、computer_use 与 browser_use 分工、桌面输出目录约定。 - config:桌面产物分类输出目录(MANUS_OUTPUT_DIR 可覆盖)、浏览器视口尺寸配置。 - tests/test_computer_use.py:接线/降级纯逻辑 + importorskip 守卫的真机截图。 健壮性修复: - app/__init__.py:预设 KMP_DUPLICATE_LIB_OK 规避 faiss/MKL 与 LLVM OpenMP 双运行时在 Windows 上触发 OMP Error #15 中止。 - llm.py:qwen3.7-plus 标记为多模态(实测可看图),并仅在确有图片时才告警。 - python_execute.py:改用 python -c 子进程执行(避开 Windows spawn 重导入 整个 app 包导致启动即超时),强制 UTF-8,默认超时 5→30s。 - str_replace_editor.py:本地目录列举改跨平台 Python 遍历,不再依赖 Windows 上不存在的 Unix find。 - file_operators.py:子进程输出 decode(errors="replace") 防中文 Windows 解码崩溃。 - gui_agent.py:修复 Qwen-VL 偶发畸形坐标 JSON;支持系统提示词覆盖。 文档:docs/batch-collect-refactor-plan.md、docs/computer-use-and-isolation.md、 README_批量采集.md。 本机一次性调试产物(probe_upload.json、repair_pawly.py,含硬编码本地路径) 已加入 .gitignore,不入库。
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
围绕"把 Temu 已发布商品批量采集到 WPS 成本核算表"的完整能力,含支撑设施与
Windows/跨平台健壮性修复。
新增能力:
健壮性修复:
文档:docs/batch-collect-refactor-plan.md、docs/computer-use-and-isolation.md、 README_批量采集.md。
本机一次性调试产物(probe_upload.json、repair_pawly.py,含硬编码本地路径) 已加入 .gitignore,不入库。
Features
Feature Docs
Influence
Result
Other