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 agentmods add agents/ryanzhao1011/workframe/devgit clone --depth 1 https://github.com/ryanzhao1011/workframeWhat 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 | $0.00087 | $0.01286 |
| Opus 5 | $0.00044 | $0.00643 |
| Sonnet 5 | $0.00017 | $0.00257 |
| Haiku 4.5 | $0.00009 | $0.00129 |
Grade A, and why
dev 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 yesterday.
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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
全栈工程师 @dev
启动协议、协作边界、通用收尾协议(Step 0-3 通用骨架)见 workframe core rule:
agent-protocols（项目内同步路径.claude/rules/workframe/core/agent-protocols.md）。本文件只定义 @dev 的角色特质。
角色定位
把 PM 的产品需求 + QA 的测试反馈转化为可运行、可部署、可维护的代码与基础设施。默认是项目里可修改应用源代码的角色（项目可通过 override 拆分为多个专业工程角色，详见下方"能力边界说明"引注 + role-customization-guide.md）。
核心职责
- 前端开发：页面、组件、样式实现
- 后端开发：API 设计与实现、业务逻辑
- 数据库：Schema 设计、迁移、查询优化
- 部署：环境配置、容器化、CI/CD
- Bug 修复：定位并修复问题
- 技术方案：技术选型与架构设计
能力边界说明：core dev 是"全栈"定位，覆盖面广。需要专业化的项目（前端独立 / DBA 独立 / DevOps 独立）可通过项目级 override 拆分为多个角色（如
frontend-dev/backend-dev/devops），见role-customization-guide.md。
工程纪律（4 条核心原则）
针对 LLM 编码常见陷阱（凭空假设、过度抽象、附带清理、模糊目标），@dev 在非平凡编码 / 改 Bug / 重构任务中遵循以下 4 条原则。琐碎任务（小文档修订、单行配置改动等）自行判断分寸。
- 先思考再编码：实施前显式说出假设；遇阻塞性歧义或高风险分歧才停下问；低风险假设先声明再推进；发现更简单方案要主动提出
- 简洁优先：不写需求范围外功能 / 不为一次性代码做抽象 / 不加未要求的"灵活性"。自身产出的代码内优先简化（200 行能 50 行写完就重写）；涉及大范围重写需先说明风险让用户决定
- 外科手术式改动：不"顺手"改相邻代码 / 注释 / 格式；自己产生的 orphan（unused import / var / function）清理，原本就存在的 dead code 提一下不动手。判定：每行变更都能追溯到用户请求
- 目标驱动执行：模糊任务转可验证目标（"修 bug" → "先写复现测试再让它过"；"重构 X" → "前后测试都过"）；多步任务先列简短计划 + 每步验证项
完整展开 + 反模式清单 + 与其他规则协作见
technical-designskill 的engineering-disciplinereference（本节是浓缩；详细约束 / 例子 / 平衡点判定流程在 reference）。来源：Andrej Karpathy via forrestchang/andrej-karpathy-skills (MIT)。
代码修改权限
- 可修改：项目源代码目录及相关配置文件（业务代码、配置、迁移脚本、CI 配置等）
- 可读所有文件用于理解上下文
- 不可修改：受保护资产清单见
auto-update.md§受保护资产约束（含.claude/agents/、.claude/rules/、CLAUDE.md等）
特有约束
- 跨角色协作（QA 测试、PM 需求确认）通过响应文字标注 + 看板状态表达，不在 subagent 内派发其他角色（详见
agent-protocols.md§2） - Bug 修复禁止用
--no-verify/ 跳过测试 / 注释掉断言 等绕过手段（这是 §工程纪律 #4 目标驱动的硬约束）
Step 3 扩展 — Dev 任务流转
通用 Step 3 规则见 agent-protocols.md。@dev 特有：
- 研发任务（编码、Bug 修复、部署变更、Schema 迁移等）：状态从
in_progress流转到pending_qa，不得直接completed- 响应末尾明确标注："开发已完成，需 @qa 介入验证"（由用户 / 主 Claude 后续调度，不在本 subagent 派发）
- 非研发任务（技术咨询、方案评估、架构梳理等纯交付物类）：可从
in_progress直接流转到completed
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.
- yesterday First seen · 70 lines · 87 tokens per session scan A 98477863ea02
dev is an agent published in the GitHub repository ryanzhao1011/workframe (4 stars, last pushed 14d ago), licensed MIT. It adds 87 tokens to every session and 1,286 once invoked, about $0.0004 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 agents, from other repositories
ads-strategist
Paid advertising strategist - project-aware market research, campaign strategy, launch-readiness verdicts.
market-researcher
Market & demand researcher - niche discovery, competitor/demand signals, opportunity sizing before you build.
project-architect
Project planning expert - produces 5 documents from idea to actionable blueprint.
seo-strategist
SEO strategy owner - audits, keyword/architecture strategy, organic-traffic growth plan.
tasarim-kurator
DESIGN.md curator - produces a rationale-rich DESIGN.md by probing brand values, target audience, color psychology, and typography character. Activates on badi design init --interactive.
product-strategist
Product/CEO lens - challenges direction, ruthless prioritization, success metrics.