Token导航 LogoToken导航TokenDH.com
开发只读github未标认证来源可访问许可证需确认审计提醒

cli-forge-validateCLI forge validate 命令行

Agent Skill

cli-forge-validate 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

768

周安装

33

GitHub Stars

公开资料未说明

下载量

269
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:cli-forge-validate(CLI forge validate 命令行)
来源仓库:https://github.com/bytelandtechnology/cli-forge
仓库路径:skills/cli-forge-validate
安装命令:
npx skills add https://github.com/bytelandtechnology/cli-forge --skill cli-forge-validate
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/bytelandtechnology/cli-forge --skill cli-forge-validate

简介

cli-forge-validate 运行权威的 47 项合规规则集,验证项目是否符合发布标准。

  • 适合在发布前或定期审计时检查结构、内容与代码约束。
  • 消费 cli-plan.yml 定义检查项,产出 validation-report.yml 报告结果。
  • 安装前应确保项目已完成 scaffold 或 takeover,否则检测可能遗漏。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

cli-forge Validate

Use this stage to verify that a generated or takeover-adopted Rust CLI Skill project complies with all structural, content, and code constraints before it proceeds to publication or whenever a fresh audit is needed.

Purpose

Run the authoritative 47-check compliance ruleset against a scaffolded, takeover-adopted, or extended project.

This stage acts as the final gatekeeper before release. It consumes the cli-plan.yml to define *what* should be checked, and produces a structured validation-report.yml detailing the results.

Canonical References

Entry Gate

#CheckSource
1project_path is known and existsRouter
2Cargo.toml is present in the target directoryFilesystem
3cli-plan.yml is present (for behavioral checks)Filesystem

Required Inputs

  • Target project_path
  • .cli-forge/cli-plan.yml for expected command/flag definitions

Workflow

  1. Read ./planning-brief.md to understand the baseline planning limits.
  2. Execute the rule set defined in ./instructions/validate.md. The baseline includes 47 checks across these categories:

- STRUCT-* (structure and pipeline-artifact ignore policy) - NAME-* (directory and package naming) - DEPS-* and META-* (dependencies and Cargo metadata) - SKILL-* (the SKILL.md contract) - BUILD-* (build, clippy, and format verification) - TREE-* (command-tree contract integrity) - HELP-* (leaf/non-leaf/--help/help behavior plus man-like formatting) - DIR-* (runtime directory documentation) - CTX-* (Active Context behavior and precedence) - ERR-* (structured error behavior) - REPL-* (REPL behavior when present) - DAEMON-* (daemon lifecycle, routing, and error behavior when present) Daemon expectations are validated through the DAEMON-* ruleset and the daemon-specific narrative overlay whenever daemon behavior is present.

  1. For each check, look at the filesystem, code, and test output.
  2. If the check requires knowledge of expected CLI behaviors (for example, flags, optional feature scope, or runtime conventions), source those expectations from cli-plan.yml.
  3. Tabulate the results. A check is either PASS, WARN (missing best practice but non-blocking), or FAIL (blocks release).
  4. Determine the final aggregate result:

- compliant (0 fails) - warning (0 fails, >= 1 warn) - non_compliant (>= 1 fail)

  1. Generate .cli-forge/validation-report.yml using the template at ./contracts/validation-report.yml.tpl. The report must snapshot the contract/receipt provenance Publish will later compare against the current .cli-forge/ baseline set.
  2. Present the validation outcome and any next-stage options. When there are 2 or 3 legal next-stage options, call the runtime's dialog-based chooser (e.g., AskUserQuestion) with those options when it is available. If no dialog-based chooser is available, or if only one legal follow-up path remains, present a numbered text menu with 1 to 3 next-stage options using the smallest valid set. When only one follow-up path remains legal, show a single 1. option for that path. Put the recommended path first and add Other: explain a different follow-up request as an escape hatch. Accept only the exact digits that correspond to the numbered options actually shown, or Other:.... Do not auto-map plain-language replies onto numbered options. If a numeric reply includes additional text, or if an Other: reply conflicts with the validation guardrails, ask for clarification before proceeding. Do not require the user to type an exact phrase to continue to Publish or a fix-up stage.

Outputs

  • .cli-forge/validation-report.yml

Exit Gate

#Check
1All 47 checks executed
2Final result aggregated (compliant/warning/non_compliant)
3validation-report.yml written

Guardrails

  • WARN results do not block publication; they highlight where the cli-forge standard advises a better path but the skill operates safely.
  • FAIL results MUST block any publication or npm distribution attempt. The workflow must return to Takeover, Scaffold, Extend, or Design to correct the issue.
  • If cli-plan.yml marks daemon behavior in_scope, validate the declared daemon contract rather than silently downgrading to older managed-daemon assumptions. If the project still implements an older surface, call that mismatch out explicitly.
  • If a pre-existing project lacks .cli-forge/cli-plan.yml, do not improvise a plan inside Validate. If .cli-forge/design-contract.yml already exists, route to Plan so the approved design wording stays authoritative. Otherwise, route to Takeover to reconstruct the missing baseline.
  • Do not publish automatically. Ensure the user has the chance to review the validation report before handing off to the Publish stage. Use the runtime's dialog-based chooser (e.g., AskUserQuestion) for the handoff when there are 2 or 3 legal next-stage options. If no dialog-based chooser is available, or if only one legal follow-up path remains, use the standardized numbered text fallback with Other: explain a different follow-up request.

Next Step

  • If compliant or warning, proceed to ../cli-forge-publish/SKILL.md.
  • If non_compliant, return to Takeover, Scaffold, Extend, or Design to fix the underlying issues.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

32.81%
按下载量换算88

Claude

32.49%
按下载量换算87

Cursor

20.74%
按下载量换算56

Gemini CLI

9.77%
按下载量换算26

安全审计

Gen Agent Trust Hub

可疑

Socket

通过

Snyk

可疑

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills