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 alisunstar/OpenSunstar --skill rd-protocolgit clone --depth 1 https://github.com/alisunstar/OpenSunstarWrote 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/alisunstar/opensunstar/rd-protocol)<a href="https://agentmods.dev/skills/alisunstar/opensunstar/rd-protocol"><img src="https://agentmods.dev/badge/skills/alisunstar/opensunstar/rd-protocol/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/alisunstar/opensunstar/rd-protocol"><img src="https://agentmods.dev/badge/skills/alisunstar/opensunstar/rd-protocol.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00061 | $0.01040 |
| Opus 5.5 | $0.00024 | $0.00416 |
| Sonnet 5.5 | $0.00012 | $0.00208 |
| Haiku 4.5 | $0.00006 | $0.00104 |
Grade A, and why
rd-protocol 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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
RD Protocol — AI 研发交付命令协议
本 skill 定义
/rd:*命令链，让外部 CLI（Claude Code/Codex/Gemini CLI 等）按协议跑通 AI 研发交付全链路。 OpenSunstar 只做协议分发与产物门禁，不持有 agent 运行时（C1 约束）。
七步闭环
verify-prd → clarify/analyze → decompose → verify-requirement → apply → validate → code-review/release → backfill
命令清单
| 命令 | 阶段 | 产物 | 人工确认 |
|---|---|---|---|
/rd:work |
路由入口 | 下一步引导（命令+推荐提示词） | 否 |
/rd:verify-prd |
输入质量门禁 | CHANGE.md | 是 |
/rd:clarify |
澄清 | clarification.md | 是 |
/rd:analyze |
分析（读 ROUTING 加载知识） | analysis.md | 否 |
/rd:decompose |
按应用拆解 | DECOMPOSITION.md + applications/{app}/requirement.md | 是 |
/rd:verify-requirement |
契约确认 | — | 是 |
/rd:apply |
代码实现 | — | 否 |
/rd:validate |
五状态对账 | IMPLEMENTATION-CHECK.md + CONTINUE-PROMPT.md | 是 |
/rd:code-review |
代码审查 | REVIEW.md | 是 |
/rd:release-plan |
发布计划 | — | 是 |
/rd:backfill |
知识回补 | KNOWLEDGE-BACKFILL.md → candidates | 是 |
/rd:extract-routing |
源码锚点提取适配器（外置） | knowledge/candidate/anchors-*.md 草稿 | 是 |
/rd:backfill-auto |
回补生成器适配器（外置） | wiki 候选（engine=rd-backfill-auto） | 是 |
通用协议规则
每条命令遵循以下约定：
- 先读规则：执行前先读 knowledge/ROUTING.md 定位相关知识，按渐进式加载层级读取
- 何时澄清：遇到歧义/缺失/冲突时暂停，产出 clarification，等人工确认后继续
- 何时停止：门禁未通过（verify-prd/verify-requirement/validate）时停止，不自动绕过
- 产物写哪：所有产物落盘到
.specs/<change-id>/ - 需人工确认：上表"人工确认=是"的命令，产出后暂停等确认
- fail-fast：能在早期阶段暴露的问题不拖到后期
与 rd-loop preset 的关系
- 本协议包是命令协议层（定义 /rd:* 怎么工作）
- rd-loop preset 是 RD 过程资产层（定义阶段/工件/门禁）
- knowledge recipe 是知识资产层（定义五层目录/ROUTING/回补）
- 三者共同构成三层资产 MVP 闭环
分发方式
rd-protocol 作为独立可发布的 skill 包分发（B2）：
- 发布物：
distrib/rd-protocol-<version>.zip（node scripts/package-rd-protocol.mjs生成）； - 安装：经 OpenSunstar Skills 管理界面「从 ZIP 安装」/仓库安装，或手工将本目录放入
~/.agents/skills/rd-protocol/； - 安装后经 skills SSOT（
~/.OpenSunstar/skills/）同步到各目标 CLI，并由 managed 标记 + agent-context 桥向 AGENTS.md 注入读取指引。
约束
- 本协议是 Markdown 资产，不含可执行代码（C2 约束）
- 命令由外部 CLI 执行，OpenSunstar 不持有运行时（C1 约束）
- 不依赖额外基础设施，纯文件协议（C3 约束）
What ships with it
14 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.
- _meta.json 770 B
- commands/analyze.md 521 B
- commands/apply.md 454 B
- commands/backfill-auto.md 1.8 KB
- commands/backfill.md 1.5 KB
- commands/clarify.md 335 B
- commands/code-review.md 521 B
- commands/decompose.md 950 B
- commands/extract-routing.md 2.0 KB
- commands/release-plan.md 382 B
- commands/validate.md 1.3 KB
- commands/verify-prd.md 1.1 KB
- commands/verify-requirement.md 518 B
- commands/work.md 1.7 KB
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 · 69 lines · 61 tokens per session scan A 7b4712e8cce4
rd-protocol is a skill published in the GitHub repository alisunstar/OpenSunstar (31 stars, last pushed 17d ago), licensed Apache-2.0. It adds 61 tokens to every session and 1,040 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
pii-redaction
Redact or remove PII (SSNs, EINs, names, addresses, phones, emails, account numbers) from documents while keeping the file in its original format and style. Use when asked to redact, mask, anonymize, scrub, sanitize, or de-identify a document.
financial-analysis
Financial statement analysis methodology covering the income statement, balance sheet, and cash flow statement, plus ratio analysis and trend interpretation for assessing a company's health.
research
Deep, multi-source research and investigation methodology for gathering, verifying, and synthesizing information from primary and secondary sources into a cited report.
supericons-icon-search
Use Supericons MCP when the user needs to find, preview, compare, retrieve, or recommend SVG icons for software UI.
cw-gates
Use before claiming any Codewhale change is done, green, or ready to land: the focused-to-broad verification ladder, the budget checks CI enforces, and the rules for what counts as a passing test.
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.