feat(sdd): 同步上游 v6 子智能体驱动开发重写(合并审查者 + scripts/)#108
Merged
Conversation
对齐 obra/superpowers v6.0.0+(当前 v6.1.1)对 subagent-driven-development 的重写: - 双审查者(规格 + 代码质量)合并为单个 task-reviewer-prompt.md,一次读 diff 出两个结论 - 新增 scripts/:review-package(生成审查包)、sdd-workspace(工作树临时区)、task-brief(抽取单任务简报) - SKILL.md 补齐 v6 新增章节:起飞前计划审查、处理审查者⚠️ 事项、构造审查者提示词、文件交接、持久化进度(ledger) - implementer-prompt.md 改为简报文件 + 报告文件契约、model 必填、TDD RED/GREEN 证据 - 删除 v5 遗留的 spec-reviewer-prompt.md / code-quality-reviewer-prompt.md 中文 fork 适配:task-brief 的 awk 同时匹配英文 "Task N" 与中文 "任务 N" 标题, 使 writing-plans 产出的中文计划也能抽取任务。review-package / sdd-workspace 与上游逐字节一致。 关联 #89 #19
Closed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
你要解决什么问题?
用户在 #89、以及同步追踪 #19 中反馈:superpowers-zh 的
subagent-driven-development仍停留在上游 v5 的双审查者架构,没有跟上上游 v6.0.0 起对该 skill 的重写(当前上游 v6.1.1)。具体表现:本地
SKILL.md还画着「先分派规格审查子智能体 → 再分派代码质量审查子智能体」的旧流程;目录里是spec-reviewer-prompt.md+code-quality-reviewer-prompt.md两个模板;缺少上游 v6 新增的scripts/(review-package / sdd-workspace / task-brief)与「文件交接、持久化进度账本、构造审查者提示词」等章节。即中文用户拿到的是过时一个大版本的 SDD 工作流。这个 PR 做了什么改变?
把
subagent-driven-development对齐到上游 v6.1.1:双审查者合并为单个task-reviewer-prompt.md,新增scripts/三件套,SKILL.md/implementer-prompt.md补齐 v6 新增契约(简报文件 + 报告文件、model 必填、TDD RED/GREEN 证据、进度账本),删除 v5 遗留的两个 reviewer 模板。这个改变适合放在核心库中吗?
适合。这是对上游 core skill 的中文同步翻译,通用、与项目类型无关,属于本 fork「完整翻译上游 skill」的既定范围。未引入任何第三方依赖。
你考虑了哪些替代方案?
task-brief的 awk 只认英文## Task N标题,而本 fork 的writing-plans产出的是### 任务 N:…,照抄会导致中文计划抽取不到任务。故对task-brief的正则做了最小适配,同时匹配Task与任务(review-package/sdd-workspace无语言依赖,与上游逐字节一致)。这个 PR 是否包含多个不相关的改变?
否。全部围绕
subagent-driven-development一个 skill 的 v6 同步。已有的 PR
测试环境
评估
如实说明:本 PR 是对上游「已经过上游校准」的 v6 内容做中文翻译,不是新的行为设计。 已完成的验证:
sdd-workspace正确建自忽略工作区(git status 干净);review-package正确按BASE..HEAD生成含提交列表/stat/带上下文 diff 的审查包(验证多提交场景不被HEAD~1截断);task-brief对中文「任务 N」、英文「Task N」两类计划均正确抽取,且「任务 1」不会误吞「任务 10」,缺失任务正确以退出码 3 报错。bash scripts/audit.sh:106 PASS / 0 FAIL;SDD 的 SKILL.md 标题结构与上游对齐(未触发结构漂移告警)。尚未做: 未跑
tests/subagent-driven-dev/run-test.sh的真实多智能体集成 eval(该 eval 需实拉 agent 会话、有 API 成本)。建议合并前由维护者补跑一次 go-fractals / svelte-todo 集成 eval 作为把关。严格性
人工审核
task-brief的中文正则适配与三个 prompt 译文)。