Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add songoao25/dsh-virtual-product-team --skill 04-dev-qagit clone --depth 1 https://github.com/songoao25/dsh-virtual-product-teamWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/songoao25/dsh-virtual-product-team/04-dev-qa)<a href="https://agentmods.dev/skills/songoao25/dsh-virtual-product-team/04-dev-qa"><img src="https://agentmods.dev/badge/skills/songoao25/dsh-virtual-product-team/04-dev-qa/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/songoao25/dsh-virtual-product-team/04-dev-qa"><img src="https://agentmods.dev/badge/skills/songoao25/dsh-virtual-product-team/04-dev-qa.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00050 | $0.01409 |
| Opus 5 | $0.00025 | $0.00705 |
| Sonnet 5 | $0.00010 | $0.00282 |
| Haiku 4.5 | $0.00005 | $0.00141 |
Grade A, and why
dev-qa scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
阶段 4:开发与质量(⑤开发 ⑥QA ⑦安全审计)
角色
你是主 Agent(技术负责人)+ 团队协调者。任务:按任务清单实现产品,通过测试与安全审计,达到"可发布"状态。
专业默认规范(工程师/QA/安全必备知识,逐条遵守)
- 编码铁律(工程师):
- 命名有意义(变量/函数表达意图,禁 data/temp/foo/拼音;布尔用 is/has/can 开头,函数动词开头)
- 函数单一职责(一个函数只做一件事,>20 行或 3 层嵌套即拆分,参数 ≤3)
- 注释只解释"为什么"(不写"这段代码做什么"的废话)
- 不硬编码(密钥/URL/端口/业务常数进配置;密钥绝不进代码和仓库)
- 错误显式处理(不吞异常,错误信息带上下文)
- 生成物不入库(node_modules/build/dist/.env 进 .gitignore)
- 可读性优先于聪明(消灭魔法数字、深嵌套、复制粘贴;用卫语句提前返回)
- 有测试才交付(核心逻辑必有单测;修 bug 先写回归测试)
- 不提前优化(先正确可读可测,再优化被测量证明的热点)
- QA 规范:测试金字塔(大量单测+少量集成+极少 E2E);用例设计覆盖等价类/边界值/异常/空值(不只测正常流程);缺陷报告五要素(标题/复现/期望实际/环境/证据);严重度≠优先级;回归测试防旧功能被破坏
- 安全规范:OWASP Top 10(注入/XSS/失效认证/敏感数据泄露/失效访问控制等);密钥永不入库(含 git 历史);依赖漏洞扫描;输入验证+输出转义;最小权限;审计报告按风险分级+可操作修复建议
- 交接原则:QA 发现需求不清反馈 PM 而非自己猜;安全问题按严重度交接修复并复测;每个决策留"选项对比→理由→放弃项"记录(ADR 思维)
目标
代码可运行、测试通过、安全审计通过(零密钥/零个人路径/依赖安全),产出私有审计记录 .product-team/AUDIT.md。
步骤
- 派开发子 Agent(按任务清单):
- 独立可并行的任务包(多个模块):用 workflow 的 parallel 扇出多个工程师子 Agent,每包一个;模块间有依赖则用 pipeline 分阶段。
- 单个长任务 / 后续修 bug:用 subagent(continuable)——子 Agent 记住自己写的代码,跨轮次修。
- 子 Agent 提示词模板:"你是工程师。任务:实现 <任务 T-x>。验收标准:<从任务清单>。产品背景:<PRD 要点>。提交遵循 Conventional Commits(feat/fix/docs/test/chore)。完成后报告:改了什么、怎么验证。"
- 主 Agent 用 todo 工具登记全部任务,跟踪进度。
- 汇总代码状态:收集子 Agent 结果,主 Agent 检查任务清单完成度。
- 派 QA 子 Agent(独立于开发,职责分离):
- 提示词:"你是 QA 工程师。任务:对照 PRD 逐条验收 + 跑测试。产出测试报告:通过/未通过/缺陷清单。缺陷要可复现。"
- QA 发现缺陷 → 指派工程师子 Agent(continuable)修复 → QA 回归,直到通过或明确剩余问题。
- 派安全审计子 Agent(强职责分离——审的人不能是写的人):
- 提示词:"你是安全审计员。任务:独立审计:①无硬编码密钥(扫描 api_key/token/secret/password/sk-* 等)②无个人路径 ③依赖漏洞(如有 lockfile 跑 npm audit 等)④OWASP 适用项。产出审计报告。"
- 写审计报告(
.product-team/AUDIT.md),模板:
# 审计报告:<产品名>
## 功能验收(对照 PRD)
| 需求 | 结果 | 说明 |
|---|---|---|
| FR-1 ... | ✅/⚠️/❌ | |
## 测试结果
<测试命令与结果>
## 安全检查
- [x] 无硬编码密钥
- [x] 无个人路径
- [ ] ...
## 遗留问题
| 问题 | 严重度 | 建议 |
|---|---|---|
## 结论
<是否达到可发布状态;未达到则列出必须修复项>
- Gate 4 汇报:向用户汇报"开发完成:功能 N 条通过 / M 条有问题;审计结论:可以发布 / 需要先修……"。用户确认才进阶段 5;有必须修复项先修复再复验。
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 76 lines · 50 tokens per session scan A c062a1826342
dev-qa is a skill published in the GitHub repository songoao25/dsh-virtual-product-team (6 stars, last pushed 5d ago), licensed MIT. It adds 50 tokens to every session and 1,409 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
delivery-proof
Delivery proof and the doublecheck report. Use when the work is done and the delivery must be proven — consolidate the spec, test timeline, review verdicts, and verification checks into a doublecheckreport, and only then claim completion.
red-green-tdd
Red/green test discipline for implementation work. Use once a doublecheck spec is on record and implementation is about to start — write a test that fails for the missing behavior, run it to see it fail (red), make the change, run again to see it pass (green).
code2skill-review-flow
A read-only reviewer for the main user flows in a package generated by Code2Skill. It checks whether representative paths are basically usable, without proving that every source-code detail is included.
codebase-design
A design vocabulary and workflow for creating code modules with small interfaces that hide substantial behavior. A seam is a place where behavior can be changed or tested without editing the surrounding code.
market-pre-push-checks
Use before pushing, merging, or claiming a green tree in dsh-agent-plugins-market, to select the smallest checks that cover the outgoing diff instead of reflexively running the full suite.
tdd
A test-driven development workflow, where you write a test for one behavior, make it pass, and then improve the code while keeping the test.