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 xiaozhi86/qamaster --skill requirement-reviewgit clone --depth 1 https://github.com/xiaozhi86/qamasterWrote 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/xiaozhi86/qamaster/requirement-review)<a href="https://agentmods.dev/skills/xiaozhi86/qamaster/requirement-review"><img src="https://agentmods.dev/badge/skills/xiaozhi86/qamaster/requirement-review/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/xiaozhi86/qamaster/requirement-review"><img src="https://agentmods.dev/badge/skills/xiaozhi86/qamaster/requirement-review.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 | $0.00014 | $0.09613 |
| Opus 5 | $0.00007 | $0.04806 |
| Sonnet 5 | $0.00003 | $0.01923 |
| Haiku 4.5 | $0.00001 | $0.00961 |
Grade A, and why
requirement-review 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 2d 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 — 916 lines — stays where its author put it; the contents beside it link to each section on GitHub.
★ Runtime 控制协议(最高优先级·模型无关·必读)
流程控制权不在模型,在 Runtime。 本 skill 的业务规范(本文件)由 Python 状态机
runtime/qamaster_runtime.py驱动执行;无论底层是什么模型,0→1→…→7 阶段顺序由状态机裁决,模型只负责当前阶段的思考与产物。
你的角色
你是 LLM Worker:只在 Runtime 颁发的【RUNTIME CONTRACT 契约卡】范围内思考与产出。你无权:决定下一阶段、宣布阶段完成、跳过人工门禁、修改流程状态。
入口协议(bootstrap → start,单步不变·模型无关)
/requirement-review命令文件内部链式跑两步,用户无感;模型只接收 Runtime 颁发的契约卡。
- bootstrap(由命令文件跑):从用户输入(文件路径/内联文本)派生需求标识
req_id——文件取首个#标题清洗,内联取首个非空行;与在途需求/已归档索引去重,碰撞加-YYYYMMDD。不创建状态(幂等可重跑);检测到进行中状态则输出RESUME,start走 resume 分支不重建。 - start --req-id (由命令文件跑):req_id 必需且恒非空;状态落
.qamaster/requirement-review/<req_id>/state.json;启动或断点续跑;输出 Phase 0 契约卡。
模型不派生 req_id:Phase 0 起所有产出物文件名直接用 state.req_id(来自 bootstrap),不再在阶段内派生 id——消除"先有鸡还是先有蛋"。重跑 /requirement-review 同一在途需求:bootstrap 输出 RESUME → start 续跑,断点不丢。
每轮执行循环(强制)
读契约卡(start/next/status 的输出)
→ 按 ALLOWED 执行当前阶段,产出 PRODUCES
→ 运行 python "runtime/qamaster_runtime.py" gate --workflow requirement-review --req-id <id>
PASS → 运行 next --workflow requirement-review --req-id <id> 取下一阶段契约卡
FAIL → 按修复指令原地修复,重跑 gate(禁止跳阶段)
→ 人工门(Phase 0 澄清基线门 + Phase 4 用户确认):
Phase 0 先落盘 Baseline_<id>.md(含 BASELINE_STATUS: clear|ambiguous),
基线 clear → gate 自动 PASS(零交互);ambiguous → WAIT,澄清 ≤3 题后重写 clear 再 gate;
Phase 4 输出确认请求后停止等待用户;确认用 confirm --workflow requirement-review --req-id <id>
Phase 4 反馈问题用 fail --to <阶段> --workflow requirement-review --req-id <id> --reason "..."
铁律
- 状态以 Runtime 为准:每次接到用户新消息,先运行
python "runtime/qamaster_runtime.py" status --workflow requirement-review --req-id <id>恢复权威状态,禁止凭对话记忆推断"现在该哪一步"。 - 门禁以机器判定为准:
gate的 PASS/FAIL 由确定性检查给出;禁止模型自证"已通过"(声明≠核实)。 - MANIFEST 由 Runtime 维护:
requirement-review-out/MANIFEST.md是多需求共享索引,由 Runtime 在 gate PASS 时自动维护(Phase 0add/ Phase 1update评审问题清单 / Phase 5update最终需求文档 / Phase 7complete)。模型禁止 Write/Edit MANIFEST.md。失步时执行python runtime/qamaster_runtime.py manifest reconcile重建。 - 业务规范不变:Runtime 只做流程控制;专家 Agent 评审标准、输入协议、输出协议等全部业务规则仍以本文件为唯一细则来源。流程由 Runtime 严格控制、与模型无关。
What ships with it
6 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago Changed · +18 lines 05139aa1e8aa
- 10d ago First seen · 898 lines · 14 tokens per session scan A 58ee01e7f56c
requirement-review is a skill published in the GitHub repository xiaozhi86/qamaster (5 stars, last pushed 3d ago), licensed MIT. It adds 14 tokens to every session and 9,613 once invoked, about $0.0001 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
Book2Skill
Convert one or more TXT, Markdown, DOCX, or PDF documents into a reusable skill zip backed by normalized Markdown, extracted images, and a grounded JSONL knowledge index. Invoke this skill before inspecting task files, then execute its workflow directly without listing directories.
browserwing-executor
Control browser automation through HTTP API. Supports page navigation, element interaction (click, type, select), data extraction, accessibility snapshot analysis, screenshot, JavaScript execution, and batch operations.
dev-browser
Browser automation with persistent page state. Use when users ask to navigate websites, fill forms, take screenshots, extract web data, test web apps, or automate browser workflows. Trigger phrases include "go to [url]", "click on", "fill out the form", "take a screenshot", "scrape", "automate", "test the website"…
ask-user-question
Ask users questions via the UI. Use when you need clarification, user preferences, or confirmation before proceeding. The user CANNOT see CLI output - this tool is the ONLY way to communicate with them.
docx
Read, edit, or create Microsoft Word .docx files. Trigger this skill whenever the user mentions a Word document, .docx file, contract, report, brief, memo, or asks to extract text, modify an existing doc, generate one from a brief, or audit tracked changes. Three execution paths: text-and-structure extraction…
pdf-toolkit
Structured .pdf operations: extract text/tables, merge pages from multiple PDFs, split a PDF by page ranges, fill PDF form fields, and generate fresh PDFs from JSON. Trigger when the user wants programmatic PDF work without natural-language rewriting — examples: pull tables from a report, combine three PDFs, extract…