Pull Request已成功合入, 合并人@CANN-robot
（感谢 iuyi 的贡献）
🤖 正在生成合并请求摘要，请稍候…
🤖 AI 代码检视正在进行中，请稍候…
Hi @iuyi, welcome to submitting your first PR to cannbot-skills!
PR Merge Steps
1. CLA Signing
If the current PR label includes cann-cla/yes, it means you have signed the CLA and can proceed to the next step. If the label includes cann-cla/no, please sign the CLA first. If you have any questions, please refer to the FAQ.
2. CI Check
Please comment /compile to trigger the CI pipeline check. If the CI run is successful, the PR will be tagged with ci-pipeline-passed and you can proceed to the next step. If the CI run fails, the PR will be tagged with ci-pipeline-failed, please check the CI logs to fix the issues in the PR. If you have any questions, please refer to the FAQ.
3. Code Review
After CI passes, please refer to the PR Approval Progress and proactively @ the committers in the table to review the code. After approval, committers will comment /lgtm and /approve. Once the lgtm and approved labels are successfully added, the PR will be merged automatically.
Thanks for your pull-request.
The full list of commands accepted by me can be found at here.
You can get sig-info at here.
You can self-configure the PR merge rules for this repository. For more details, please refer to Here.
For more, you also can visit HICANN.
PR Approval Progress
✅ Congratulations! All modules have met the lgtm and approve requirements.
Module Approval Details
| module | lgtm status | approve status |
|---|---|---|
| repo-cann/cannbot-skills | ✅ 汤平川, xutianze (2/2) | ✅ 汤平川, xutianze (2/1) |
💡 Tip:
- Committer can comment
/approveor/lgtm- Commenting
/approveimplies both code review (lgtm) and intent to merge (approve)
CLA Signature Pass
iuyi, thanks for your pull request. All authors of the commits have signed the CLA. 👍
⚠️ 本次变更过大（40 个可审文件、5303 行，预估需 5 轮分片审查），已超出 AI 代码评审的处理范围，本次跳过。建议拆分为更小的 PR 以获得有效评审。
⚠️ 本次变更过大（40 个可审文件、5303 行，预估需 5 轮分片审查），已超出 AI 代码评审的处理范围，本次跳过。建议拆分为更小的 PR 以获得有效评审。
compile
iuyi
/approve
流水线任务触发成功
任务链接 [2abf246b418b40d9a3d7bef761b70b57][流水线指导]
| 任务名称 | 状态 | 日志 | 下载链接 |
|---|---|---|---|
| codecheck | ✅ SUCCESS | >>>>> | |
| SCA | ✅ SUCCESS | >>>>> | |
| antipoison | ✅ SUCCESS | >>>>> | |
| Check_Pr | ✅ SUCCESS | >>>>> | |
| UT_Test | ✅ SUCCESS | >>>>> | |
| codecheck_style | ✅ SUCCESS | >>>>> | |
| pre-commit | ✅ SUCCESS | ||
| pre_comment | ✅ SUCCESS | >>>>> | |
| UT_Test_AI_review | ✅ SUCCESS | >>>>> | |
| ST_Test_A2 | ✅ SUCCESS | >>>>> | |
| ST_Test_A5 | ✅ SUCCESS | >>>>> |
[2026-09-01 11:37:09] CI执行结束
/lgtm
/approve
描述
本次重构将 ops-direct-invoke 插件的工作流实现从「静态 agents + workflows 目录」模型改为 skill 驱动的多角色编排模型，借鉴面向对象设计思想，将通用工作流沉淀为 CANNBot 维护的「基类」，各算子仓作为「子类」继承编排流程、仅覆写与自身耦合的领域知识，达到一套核心流程服务所有算子仓、消除重复建设的目的。
主要变更
workflows/静态目录，新增 16 个子 skill：ops-direct-invoke-workflow：统一流程编排（阶段 0-7、CP0-CP5 检查点、任务回退、静默模式）ops-direct-invoke-workflow-maintain：维护红线与 SOLID 设计约束看护plugin-experience-summary/plugin-perf-iteration/plugin-pr-submit：可插拔流程插件，frontmatter 声明挂载点、init 注册到.cannbot/settings.json，自闭环可单独触发repo-build-guide/repo-coding-rules/repo-knowledge/repo-op-templates/repo-test-develop：仓库领域知识workflow-cp0~workflow-cp5：各检查点验收 skillworkflow-doc-templates/workflow-agent-permissions：文档模板与权限管理.cannbot/settings.json唯一权威（工作流模式、插件注册与启用状态、问卷状态）；新增静默模式（--mode silent完全无人值守，支持会话内「开启/关闭静默模式」切换与权限预检）。.cannbot/permissions、权限 hook、.cannbot/settings.json、依赖仓 clone 到.cannbot/；example/init.sh提供算子仓接入范例。dependency_validator.py增强 DG-11 校验，支持从agents/*.mdfrontmatter 解析 skill 白名单。关联的Issue
测试
我们在 CannBench 官方任务集上完成 53 个算子的全量实现与评测（方案名 direct-invoke@Claude deepseek-v4-flash），标准集用例通过率 98.8%、隐藏集 95.5%，总分 7383.64，平均加速比标准集 0.53、隐藏集 0.85。
文档更新
plugins-official/ops-direct-invoke/README.md：设计思想、知识分层、角色与流程概览、设计约束、静默模式说明类型标签
流水线故障排查