agent-dx-cli-cn
Skill14 starsChanged 22 days agoApache-2.0View on GitHub
设计、审查和重构面向 AI Agent 的 CLI。检查可预测性、纵深防御和上下文效率,并用实际证据评分。用于开发 Agent DX CLI 或检视现有 CLI。
1 other project uses a version of this.
Adopted by
Most recently taken up by
- ontology-of-everything/concept-skillsword for word, first seen Nov 15, 2026
Other versions1 project uses a version of this.
Word for word · 1
- ontology-of-everything/concept-skills★ 14100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository22 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- agent-dx-cliSkill
- architect-4cSkill
- architect-4rSkill
- architect-frontend-feature-firstSkill
- architect-frontend-fsdSkill
- concept-designSkill
- concept-modelSkill
- data-knowledge-architectSkill
- personal-knowledge-architectSkill
- read-it-reframe-it-own-itSkill
- simplify-javaSkill
- architect-4c-cnSkill
- architect-4r-cnSkill
- architect-frontend-feature-first-cnSkill
- architect-frontend-fsd-cnSkill
- concept-design-cnSkill
- concept-model-cnSkill
- data-knowledge-architect-cnSkill
- personal-knowledge-architect-cnSkill
- read-it-reframe-it-own-it-cnSkill
- simplify-java-cnSkill
The file
Files
Agent DX CLI
设计、审查和重构面向 Agent 的 CLI。保留用户范围、技术栈和已有授权。 人类便利功能与 Agent 契约可以共存;只引入任务需要的机制。
工作流程
- 检查代码、当前 help/schema 和实际输出,记录 CLI 产物与版本。
- 阅读原则、理由与正反例,评估全部适用检查点,区分缺失、未验证和已验证。
- 实施已授权修改,优先修复验收阻断项。复用成熟库,CLI、MCP 与批量入口共用校验和执行路径。
- 在已授权的隔离夹具中验证正反例,报告修改、逐项证据、评分和未解决的限制。
事实和授权足够时继续执行。仅在歧义改变目标或后果时澄清。审查请求只授权审查。
原则
- **可预测性:**从输入到结果保留类型化意图,通过可查询契约减少猜测。
- **纵深防御:**分别校验输入、落实当前权限、核对执行结果;一种保护不能替代另一种。
- **上下文效率:**返回有界数据,按任务加载知识;每个额外字段或指令都应有实际价值。
执行边界
- 默认非交互执行。可选交互需显式选择,并要求 TTY。
- 将 Agent 输入视为不可信数据,在共用执行器中按类型和字段用途落实约束,先校验再产生业务副作用。
- 服务端内容即使经过净化仍按不可信数据处理,保留来源。
- 破坏性操作必须提供
--dry-run。声明预演的实际副作用。 - schema、预演、质量报告和确认开关不授予权限。执行时重新检查授权。
- 结果未确定时返回
unknown。先核对结果再安全重放;确认用户要求的结果后才报告完成。
可选参考
- 对齐 API/GWS 时,使用 GWS 实践,保留固定参考版本并验证项目差异。
- 编写或审查配套技能时,使用 SKILL+CLI 编写,区分四层渐进披露职责,保留有效的决策指导。 实质性修改的指导需与无技能基线对照;过时语法或误导规则可能使表现更差。
评分与验收
模式参考定义十个维度,每维两项检查,每项 0–5 分,总分 100。这是本技能的评分约定。
| 分值 | 所需证据 |
| 0 | 缺失或存在反证。 |
| 1 | 只有文档声明。 |
| 2 | 已实现但未验证。 |
| 3 | 正例通过。 |
| 4 | 相关负例通过。 |
| 5 | 目标环境端到端通过。 |
静态项:3 为实现审查,4 为消费者与负例对照,5 为可复现的契约或依赖检查。 每项评分关联证据和测试产物/版本。必要行为未验证时不能通过验收。 验收要求总分至少 90,校验和认证各至少 9。 凭据泄露、授权绕过、预演业务写入、危险重放或虚假完成,无论得分多少都阻断验收。 逐项说明 N/A;已要求但缺失的功能不能标为 N/A。报告原始分和换算分:原始分 ÷ 适用满分 × 100。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

