Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问许可证需确认审计通过

kata-linearkata Linear 搜索

Agent Skill

用于处理 Linear 项目、Issue、团队、周期和产品开发任务流。它适合让 Agent 辅助查询任务状态、整理需求队列、创建缺陷或汇总迭代进展。使用时需要确认 workspace、team、label、assignee 和状态流转规则;涉及批量创建或修改任务时,应先核对字段和目标团队,避免把草稿需求直接写入正式项目。

总安装

783

周安装

32

GitHub Stars

1

下载量

253
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:kata-linear(kata Linear 搜索)
来源仓库:https://github.com/gannonh/skills
仓库路径:skills/kata-linear
安装命令:
npx skills add https://github.com/gannonh/skills --skill kata-linear
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/gannonh/skills --skill kata-linear

简介

kata-linear 用于处理 Linear 项目、Issue、团队和产品开发任务流。

  • 适合查询任务状态、整理需求队列、创建缺陷或汇总迭代进展。
  • 使用时需确认 workspace、team、label 和状态流转规则。
  • 批量创建任务时应核对字段和目标团队,避免误写正式项目。
  • 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

Linear Ticket Lifecycle

This skill wraps the general /linear skill with structured start/next/end workflows. Always invoke the /linear skill for the actual MCP calls when using this skill.

Determining the Mode

Parse the user's request to determine which workflow to run:

User saysMode
"start KAT-N", "pick up KAT-N", "implement KAT-N", "work on KAT-N"Start
"finish KAT-N", "complete KAT-N", "done with KAT-N", "close KAT-N"Complete
"what's next", "what should I work on", "next ticket", "next issue"Next

If ambiguous, ask.

Identifying the Project

  1. Check the current working directory's AGENTS.md for a Linear project reference.
  2. If not found, call list_projects and ask the user which project to use.
  3. Cache the project name for the remainder of the session and update AGENTS.md for future reference.

Mode: Next

Find the next actionable issue.

  1. Query list_issues for the project with state Todo.
  2. If results exist, present them. The first Todo issue is the recommended next pick.
  3. If no Todo issues, resolve from blocking relations: a. Query list_issues for the project with state Backlog. b. For each Backlog issue, call get_issue with includeRelations: true. c. Find issues whose blockedBy entries are all Done (or have no blockers).
  4. If multiple candidates, determine if they can be run concurrently in isolated worktrees in two ways: a. Re-check Linear tickets (scope + blockers); then b. Inspect the current code boundaries they would touch to see if there’s hidden coupling.
  5. Present unblocked issues as candidates, including concurrency guidance/risks.
  6. If the project has Linear documents (execution model, workflow contract), fetch them with list_documents and get_document to understand pillar/phase ordering.

Mode: Start

Step 1 — Validate the issue

  1. Call get_issue for the requested issue with includeRelations: true.
  2. Check every entry in blockedBy. For each blocker, confirm its status is Done.
  3. If any blocker is not Done, stop and report which blockers remain open.

Step 2 — Move to In Progress

  1. Call update_issue to set state to In Progress.

Step 3 — Load context

  1. Read the issue description for references to specs, mocks, docs, or design files.
  2. Read Linear documents (execution model, workflow contract) for project-specific guidance.
  3. Read relevant spec files, mock images, or design references found in steps 1 and 2.
  4. Check existing source code in the areas the issue will touch.

Step 4 — Create feature branch

  1. Use the gitBranchName field from the issue response as the branch name.
  2. Create the branch from the main branch.

Step 5 — Summarize

Present to the user:

  • Issue title and acceptance criteria
  • Blocker status (all clear)
  • Context loaded (specs, mocks, relevant code)
  • Branch name created
  • Any project-specific workflow reminders from AGENTS.md (e.g., TDD mandate)

Mode: Complete

Step 1 — Gather evidence

Gather evidence:

  • PR link or branch with changes
  • Test results (unit and/or E2E)
  • Screenshots/video or spec-state references demonstrating acceptance criteria

If a PR already exists on the current branch, detect it with gh pr view.

Step 2 — Validate completion gate

Check if the project's CLAUDE.md or Linear workflow contract defines a hard gate. Common gates:

  • Referenced spec states/interactions are verified
  • Evidence links are attached (tests, screenshots, or traceable PR notes)
  • Gap analysis items are either closed or split into follow-up issues
  • Acceptance criteria are all marked as verified in the issue checklist
  • Definition of Done checklist is completed

If evidence is insufficient, list what's missing and stop.

Step 3 — Attach evidence to the issue & PR

  1. Call create_comment on the issue with a structured evidence summary: ## Completion Evidence - PR: [link] - Tests: [pass/fail summary] - Acceptance coverage: [which criteria verified] - Screenshots/video: [if applicable]
  2. Add the same summary as a comment on the PR.

Step 4 — Promote next in chain

  1. Call get_issue with includeRelations: true on the completed issue.
  2. For each issue in the blocks list: a. Call get_issue with includeRelations: true on that downstream issue. b. Check if ALL of its blockedBy entries are now Done. c. If yes, call update_issue to move it to Todo. d. Report which issue was promoted.

(Note: the current ticket will automatically move to Done when the PR is merged).

Step 5 — Summarize

Present:

  • Issue marked Done with evidence link
  • Which downstream issue(s) were promoted to Todo
  • Suggested next action

Important Reminders

  • Always pass includeRelations: true when calling get_issue to see blocking dependencies.
  • Always reference the attached media as the source of truth for design specs and mocks.

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

32.46%
按下载量换算82

Claude

31.91%
按下载量换算81

Cursor

20.24%
按下载量换算51

Gemini CLI

8.6%
按下载量换算22

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

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