Token导航 LogoToken导航TokenDH.com
开发操作浏览器github未标认证来源可访问许可证需确认审计通过

biome-enforcement生物群落执法

Agent Skill

biome-enforcement 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

563

周安装

23

GitHub Stars

1

下载量

182
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:biome-enforcement(生物群落执法)
来源仓库:https://github.com/mwillbanks/agent-skills
仓库路径:skills/biome-enforcement
安装命令:
npx skills add https://github.com/mwillbanks/agent-skills --skill biome-enforcement
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/mwillbanks/agent-skills --skill biome-enforcement

简介

定义 Agent 如何以 Biome 为代码规范强制执行工具进行操作。

  • 适合在决定哪些任务需 Biome 介入、解释诊断信息或处理修复后验证时使用。
  • 强调 Biome 作为最终修复与强制关卡的角色,保留构建与测试回验。
  • 安装需确认权限范围和维护状态,可能涉及联网、命令执行或文件读写操作。
  • biome-enforcement 属于开发类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Biome Enforcement

This skill governs how agents should work with Biome as a repository enforcement tool.

It still preserves the core rule from the prior skill:

  • Biome is primarily a final remediation and enforcement pass
  • the required final command remains fixed
  • the remediation loop remains mandatory
  • post-Biome build and test re-validation still occurs only when Biome changed files

This skill is broader than a single final step. Agents should also use it when they need to:

  • reason about whether a task is Biome-relevant
  • interpret Biome JSON diagnostics
  • decide whether generated output belongs in ignore rules
  • update Biome configuration safely
  • distinguish source issues from tool-noise issues before the final pass

Operational details for how Biome works are defined in:

references/biome.md

Agents must consult that reference when executing this skill.


Concern Ownership

This skill owns Biome-specific execution policy.

It takes precedence whenever the task involves:

  • Biome command selection
  • Biome diagnostic interpretation
  • Biome remediation sequencing
  • Biome ignore-path decisions
  • post-Biome validation rules

Repository-wide standards skills may still govern adjacent concerns, but they should defer to this skill for Biome behavior.


When This Skill Must Be Used

Activate this skill whenever a task:

  • modifies source code
  • modifies tests
  • adds or removes files in Biome-managed areas
  • changes biome.json, related workspace config, or scripts that affect Biome usage
  • creates new directories or generated output that may need ignore handling
  • requires interpretation of Biome diagnostics or JSON reporter output
  • completes an implementation that must be validated before handoff

In practice, this means most coding tasks that touch Biome-managed files should consult this skill, even if Biome itself is only executed at the end.


Execution Timing

Biome must run as the final enforcement routine after implementation and normal validation are already complete.

Agents may consult this skill earlier for planning, config awareness, or ignore-path triage, but they should not turn Biome into a continuous edit-loop tool.

Correct order:

  1. Complete implementation
  2. Run repository-native validation before Biome such as build, typecheck, and tests
  3. Run the Biome remediation loop
  4. If Biome modified files:

- re-run the same build and test validation that was already used

  1. Declare the task complete

Required Final Command

Always run Biome using the following command:

biome check --write --unsafe --changed --no-errors-on-unmatched --files-ignore-unknown=true --reporter=json

Requirements:

  • JSON reporter must always be used
  • the command must run against changed files
  • autofix must be enabled
  • this exact invocation remains the required final enforcement command

Do not substitute alternative reporters or command structures.


Broader Biome Responsibilities

This skill is not limited to "run one command at the end."

Agents should also apply it when working with:

  • Biome config updates
  • Biome ignore rules
  • parser or config failures surfaced by Biome
  • lint-rule remediation strategy
  • repository changes that expand or shrink the set of files Biome should own

These activities support the final enforcement pass. They do not replace it.


Remediation Loop

If Biome reports errors:

  1. Parse the JSON output
  2. Convert diagnostics into a remediation task list
  3. Resolve the issues
  4. Run the Biome command again

Repeat until Biome reports zero errors.

Full remediation workflow details are defined in:

references/biome.md

Ignore Path Responsibilities

If the task introduces generated artifacts or machine-produced output, the agent must determine whether those paths should be ignored by Biome.

Examples may include:

  • test artifacts
  • playwright output
  • coverage reports
  • generated snapshots
  • build output
  • tool-generated directories

Guidance for determining when ignore paths are appropriate is defined in:

references/biome.md

Agents must not add ignore rules for real source code issues.


Final Validation Requirement

Work is not complete until all of the following are true:

  • Biome reports zero errors from the required command
  • if Biome made changes, the repository-native build and test commands used earlier succeed again
  • no new Biome noise remains in the changed scope
  • any newly introduced generated output is either ignored appropriately or intentionally Biome-managed

Agent Responsibility

Agents must treat Biome as a mandatory repository enforcement mechanism.

For any coding task:

  • consult this skill when Biome-related decisions are required
  • run the required Biome command before completion
  • use the JSON reporter
  • follow the remediation loop
  • re-run build and test validation only if Biome modifies files

Agents should consult the reference document for operational details when executing this skill.

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

38.84%
按下载量换算71

Claude

32.23%
按下载量换算59

Cursor

17.12%
按下载量换算31

Gemini CLI

9.19%
按下载量换算17

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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。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