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 derricktang/pm-workflow-plugin --skill workflow-evolutiongit clone --depth 1 https://github.com/derricktang/pm-workflow-pluginWrote 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/derricktang/pm-workflow-plugin/workflow-evolution)<a href="https://agentmods.dev/skills/derricktang/pm-workflow-plugin/workflow-evolution"><img src="https://agentmods.dev/badge/skills/derricktang/pm-workflow-plugin/workflow-evolution/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/derricktang/pm-workflow-plugin/workflow-evolution"><img src="https://agentmods.dev/badge/skills/derricktang/pm-workflow-plugin/workflow-evolution.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 | $0.00049 | $0.08870 |
| Opus 5.5 | $0.00020 | $0.03548 |
| Sonnet 5.5 | $0.00010 | $0.01774 |
| Haiku 4.5 | $0.00005 | $0.00887 |
Grade B, and why
workflow-evolution scanned grade B with 1 finding 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
grep -rn "SSOT #NN\b" pm-workflow/ CLAUDE.md .claude/ How it starts
The opening of the file, as written. The whole thing — 343 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Purpose
Provide a lightweight, disciplined skill for the orchestrator (Claude Code) to handle L2 workflow maintenance tasks directly — without dispatching PM Agent + Supervisor Agent through the standard 2-round closure.
This is not a replacement for L1 product-document production (PRD/spec generation) — those still require PM/Supervisor agents. This skill applies only to L2 meta-work where the orchestrator has full context and dispatching agents adds cost without value.
Key Concepts
L1 vs L2 双层架构
| 层 | 工作性质 | 消费者 | 典型任务 | 处理方式 |
|---|---|---|---|---|
| L1 产品业务层 | 产品文档生产 | 真实业务方（产品总监 / 设计 / 开发） | PRD / spec / 阶段产物 | 派 PM Agent + Supervisor Agent |
| L2 工作流维护层 | 工作流框架自身演化 | 工作流维护者（项目维护者 + AI） | 元规范 / 校验脚本 / SSOT 双锚 / 测试 | 本 skill：编排器直做（默认）/ 派 agent（复杂大型） |
路径判定（L1 vs L2 二元）
核心判定标准：任务修改的文件归属哪一层——L1 派 Agent / L2 走本 skill。不再以文件数 / token / 是否新建脚本作为路径判定依据（这些维度仅用于"分批执行"提示，见下方）。
| 文件归属 | 路径 |
|---|---|
| L2 文件（修改任意以下路径） | 走本 skill 路径（编排器直做） |
- pm-workflow/agents/* |
|
- pm-workflow/rules/*（含 bujue-design-system/） |
|
- pm-workflow/scripts/*（含 tests/） |
|
- pm-workflow/skills/* |
|
- CLAUDE.md / agent_methodology.md / agent_parameters.md |
|
- .claude/commands/* |
|
| L1 文件 | 派 PM Agent（保留 L1 业务流程严格性） |
- outputs/（PRD / spec / 阶段产物） |
|
- process_record/state.md / process_record/tasks/（产品过程记录） |
|
| 混合任务 | L2 部分走 skill / L1 部分派 Agent（拆分执行） |
为什么 L1/L2 二元判定优于多维阈值：
- 客观：文件路径判定无主观空间，避免"50K vs 80K 算不算超阈值"的争论
- 编排器主对话天然持有 L2 上下文，PM Agent 重 Read 同样规范是浪费
- 责任清晰：L2 工作流维护是编排器的固有职责（PM/Supervisor 仅参与 L1 业务），不应通过派 Agent 转移；L2 质量保障由本 skill §L2 三道质量门负责
大型 L2 任务的分批执行模式
L2 任务规模大（参考阈值：> 200K token / 涉及 > 5 文件 / 改既有脚本结构 ≥ 30%）时，仍走本 skill 路径，但编排器分批执行：
- 拆分 atomic step 为多组（如 5 文件改动 → 拆为 file-1/2 / file-3/4 / file-5 + 测试 三批）
- 每批结束写进度文件 checkpoint（含已完成步 + 待续步）
- 批与批之间编排器自我评估上下文压力（≥ 800K token / 1M 主对话 → 提示用户考虑
/clear重启 + state.md 恢复） - 不一次性把所有文件 Read 进主对话——按批次精读
What ships with it
2 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.
- 1mo ago First seen · 343 lines · 49 tokens per session scan B c610710d4924
workflow-evolution is a skill published in the GitHub repository derricktang/pm-workflow-plugin (2 stars, last pushed 3mo ago), licensed MIT. It adds 49 tokens to every session and 8,870 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
last30Days
Resolve "last30Days" to a concrete ISO date range relative to your run time — a rolling 30-day window ending today. Returns inclusive civil dates plus exact UTC instants so you have temporal context without computing dates by hand. Read-only: no writes, no network. Use before a "last 30 days" / trailing-month task…
thisWeek
Resolve "thisWeek" to a concrete ISO date range relative to your run time — this week so far (Monday → today). Returns inclusive civil dates plus exact UTC instants so you have temporal context without computing dates by hand. Read-only: no writes, no network. Use before a week-to-date task (this week's activity…
comet-hotfix
A quick workflow for fixing an existing bug in Comet, a tool that manages structured code changes. It moves through opening the change, building, checking, and archiving it.
comet-design
A workflow guide for turning an existing software change into a detailed technical design document. It builds on earlier proposal and design files and defines implementation risks, tests, and edge cases.
comet-verify
Comet Phase 4: Verify and Close. Invoke with /comet-verify. Verify implementation matches design, handle development branch.