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 lupingluke-ai/general-ai-spec --skill prd-writergit clone --depth 1 https://github.com/lupingluke-ai/general-ai-specWrote 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/lupingluke-ai/general-ai-spec/prd-writer)<a href="https://agentmods.dev/skills/lupingluke-ai/general-ai-spec/prd-writer"><img src="https://agentmods.dev/badge/skills/lupingluke-ai/general-ai-spec/prd-writer/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/lupingluke-ai/general-ai-spec/prd-writer"><img src="https://agentmods.dev/badge/skills/lupingluke-ai/general-ai-spec/prd-writer.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 | $0.00073 | $0.04059 |
| Opus 5.5 | $0.00029 | $0.01624 |
| Sonnet 5.5 | $0.00015 | $0.00812 |
| Haiku 4.5 | $0.00007 | $0.00406 |
Grade A, and why
prd-writer 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 1mo 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 — 303 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PRD Writer
Overview
将 backlog 中的一行想法转化为结构化的产品需求文档（PRD）。读取该 backlog 条目归属的模块设计 + design-inputs，与用户对话澄清边界，生成覆盖用户故事、功能需求、验收标准的完整 PRD。
职责边界： 只负责单条 backlog → PRD。拆分工作由 module-designer 在 /design 对话中完成，本 skill 不再做粒度评估。技术规划由 change-propose，执行/审查由 change-dispatch / change-review。
Announce at start: "Running prd-writer: drafting PRD for B-NNN."
When to Use
- 用户说"帮我分析 B-NNN 的需求"
- 用户触发
/prd B-NNN - backlog 条目处于
idea阶段，模块列已填写（由module-designer产出），需要生成 PRD
Bash 命令规范
- 每条命令独立调用 — 不用
&&、||、;串联 - 管道可以用 — 如
grep -c 'B-' product/backlog.md - 并行无依赖时分开调用
输出物清单
单次运行产出（通过 governance/prd-writer/PRD-NNN PR 合入 main）：
product/prd/PRD-NNN.md（新建，含按规则生成的change-id；首次status: reviewing，用户确认 approved 后改为approved）product/backlog.md（目标 B-NNN 阶段 idea → exploring，PRD 列写入 PRD-NNN）design/modules/M-NNN-*.md（## 修订历史追加；status: planning → active当且仅当本 skill 首次将该模块的任一 backlog 推进到 exploring）design/roadmap.md的三段 AUTO 区（从 backlog + modules 全量重渲染）
Step 1 — 同步远程
git checkout main
git pull --ff-only origin main
目的：
- 读到最新 backlog 与 module 文档
- PRD-NNN 编号分配不冲突
- 不在过时基线上做产品决策
失败处理： 冲突时提示用户先解决再继续。
Step 2 — 读取 Backlog 与上下文
2.1 定位 Backlog 条目
从 product/backlog.md 读取目标 B-NNN 行，解析：
- 需求描述
- 归属模块（"模块"列的
M-NNN）— 必填字段 - type
- 阶段（应为
idea；若已exploring且 PRD 文件存在，视为增量对话）
2.2 module-ref 必填校验
若"模块"列为 — 或空：
- 历史遗留 backlog（启用本框架前创建的）→ 提示用户给这条 backlog 补模块归属：要么手工改 backlog 该列，要么通过
/design review M-NNN把它追加到模块的## 关联 Backlog - 新条目漏填 → 提示用户回
/design重新落盘（通常说明 module-designer 流程被绕过）
补齐前不生成 PRD。
2.3 必读上下文（硬性要求）
| 路径 | 用途 |
|---|---|
design/modules/<M-NNN>-*.md |
必读 — 模块边界、对外接口、技术选型。PRD 范围必须落在模块边界内 |
design/modules/<M-NNN>-*.md 的 ## 关联 Backlog 小节 |
必读 — 本 B-NNN 行尾的 [depends-on: ...]（module-designer 在对话中记录的 backlog 级依赖） |
design/inputs/** |
必读 — module 文档 design-inputs 字段列出的所有路径 |
openspec/project.md |
必读 — 技术栈约束 |
openspec/specs/ |
扫描 — 已有系统行为，避免需求冲突 |
| 相关代码 | 按需 — 评估现状时读 |
What ships with it
1 file 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.
- 1mo ago First seen · 303 lines · 73 tokens per session scan A 708d0e91f4dd
prd-writer is a skill published in the GitHub repository lupingluke-ai/general-ai-spec (5 stars, last pushed 3mo ago), licensed MIT. It adds 73 tokens to every session and 4,059 once invoked, about $0.0003 per session on Opus 5.5. 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
ai-native-audit-release
A practical method for auditing an AI product or system before release. It examines design evidence, evaluation evidence, and shadow evidence, then defines release limits and a go/no-go decision.
contract-builder
Maintain an execution contract only for an existing legacy change that requires one. New direct/planned changes do not invoke this skill when a contract is absent.
mc-zenn
Turn MulmoClaude work into a Zenn tech article (markdown) inside the workspace. On first use it sets up a Zenn project at github/zenn/ — clone an existing GitHub repo or zenn init a fresh one, idempotent and skipped when already initialized — then writes articles to github/zenn/articles/ .md tagged with the…
release-archivist
Close out a spec-superflow change with verification, summary, and archive readiness. Invoke when implementation is complete, verification is underway, or the user asks for a final wrap-up.
agt-policy-authoring
Create and validate a minimal AGT Copilot CLI policy tailored to the repository being inspected.
hermesoffice-decks
Produce presentations in HermesOffice Slides — the plandeck/generatedeck pipeline, native element editing, and the templates/decks style specs (modern, elegant, professional, keynote-speaker, minimal, tech-dark, investor-pitch, workshop). Load for any request that creates or restyles a deck.