任务创建器

Universal

ln-301-task-creator

by levnikolaevich

根据编排器提供的方案生成 implementation、refactoring、test 任务文档,校验类型规则与 DRY 风险,自动创建 Linear issue 并更新看板。

405效率与工作流未扫描2026年3月5日

安装

claude skill add --url github.com/levnikolaevich/claude-code-skills/tree/master/ln-301-task-creator

文档

Paths: File paths (shared/, references/, ../ln-*) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root.

Universal Task Creator

Worker that generates task documents and creates Linear issues for implementation, refactoring, or test tasks as instructed by orchestrators.

Purpose & Scope

  • Owns all task templates and creation logic (Linear + kanban updates)
  • Generates full task documents per type (implementation/refactoring/test)
  • Enforces type-specific hard rules (no new tests in impl, regression strategy for refactoring, risk matrix and limits for test)
  • Drops NFR bullets if supplied; only functional scope becomes tasks
  • Never decides scope itself; uses orchestrator input (plans/results)

Task Storage Mode

MANDATORY READ: Load shared/references/tools_config_guide.md and shared/references/storage_mode_detection.md

Read docs/tools_config.md (bootstrap if missing per tools_config_guide.md). Extract: task_provider = Task Management → Provider (linear | file).

Invocation (who/when)

  • ln-300-task-coordinator: CREATE (no tasks) or ADD (appendMode) for implementation tasks.
  • Orchestrators (other groups): Create refactoring or test tasks as needed.
  • Never called directly by users.

Inputs

  • Common: taskType, teamId, Story data (id/title/description with AC, Technical Notes, Context).
  • Implementation CREATE: idealPlan (1-8 tasks), guideLinks.
  • Implementation ADD: appendMode=true, newTaskDescription, guideLinks.
  • Refactoring: codeQualityIssues, refactoringPlan, affectedComponents.
  • Test: manualTestResults, testPlan (Priority ≥15, Usefulness Criteria), infra/doc/cleanup items.

Quality Criteria

MANDATORY READ: Load shared/references/creation_quality_checklist.md §Task Creation Checklist for validation criteria that ln-310 will enforce.

Workflow (concise)

  1. DRY Check (Codebase Scan): For EACH Task in plan:
    • Extract keywords: function type, component name, domain from Task description
    • Scan codebase: Grep(pattern="[keyword]", path="src/", output_mode="files_with_matches") for similar functionality
    • IF similar code found (≥70% keyword match):
      • Add ⚠️ DRY Warning section to Task description BEFORE Implementation Plan:
        markdown
        > [!WARNING]
        > **DRY Check:** Similar functionality detected in codebase
        > - Existing: src/services/auth/validateToken.ts:15-42
        > - Similarity: 85% (function name, domain match)
        > - **Recommendation:** Review existing implementation before creating new code
        >   - Option 1: Reuse existing function (import and call)
        >   - Option 2: Extend existing function with new parameters
        >   - Option 3: Justify why reimplementation needed (document in Technical Approach)
        
    • IF no duplication → Proceed without warning
    • Rationale: Prevents code duplication BEFORE implementation starts
  2. Template select: Load template based on taskType (see "Template Loading" section).
  3. Generate docs: Fill sections for each task in plan/request using provided data, guide links, and DRY warnings.
  4. Validate type rules: Stop with error if violation (see table below).
  5. Preview: Show titles/goals/estimates/AC/components, DRY warnings count, and totals.
  6. Confirmation required: Proceed only after explicit confirm.
  7. Create issues: Call Linear create_issue with parentId=Story, state=Backlog; capture URLs.
  8. Update kanban: Add under Story in Backlog with correct Epic/indent.
  9. Return summary: URLs, counts, hours, guide link count, DRY warnings count; next steps (validator/executor).

Type Rules (must pass)

taskTypeHard ruleWhat to verify
implementationNo new test creationScan text for "write/create/add tests" etc.; allow only updating existing tests
refactoringRegression strategy requiredIssues listed with severity; plan in 3 phases; regression strategy (Baseline/Verify/Failure); preserve functionality
testRisk-based plan requiredPriority ≥15 scenarios covered; each test passes Usefulness Criteria; no framework/library/DB tests

Critical Notes

  • MANDATORY: Always pass state: "Backlog" when calling create_issue. Linear defaults to team's default status (often "Postponed") if not specified.
  • DRY Check: Scan codebase for EACH Task before generation. If similar code found (≥70% keyword match) → add ⚠️ DRY Warning section with 3 options (reuse/extend/justify). Skip scan for test tasks (no implementation code).
  • Foundation-First order for implementation is preserved from orchestrator; do not reorder.
  • No code snippets; keep to approach, APIs, and pseudocode only.
  • Documentation updates must be included in Affected Components/Docs sections.
  • Language preservation: keep Story language (EN/RU) in generated tasks.

DRY Warning Examples:

markdown
Example 1: Email validation (HIGH similarity - 90%)
> [!WARNING]
> **DRY Check:** Similar functionality detected
> - Existing: src/utils/validators/email.ts:validateEmail()
> - Similarity: 90% (exact function name + domain match)
> - **Recommendation:** REUSE existing function (Option 1)

Example 2: User authentication (MEDIUM similarity - 75%)
> [!WARNING]
> **DRY Check:** Partial functionality exists
> - Existing: src/services/auth/login.ts:authenticateUser()
> - Similarity: 75% (domain match, different scope)
> - **Recommendation:** Review existing code, consider EXTEND (Option 2) or JUSTIFY new implementation (Option 3)

Example 3: No duplication (skip warning)
- No similar code found → Proceed without DRY warning

Definition of Done

  • DRY Check complete: Codebase scanned for EACH Task; similar code detected (Grep); DRY warnings added to Task descriptions if ≥70% similarity found.
  • Context check complete (existing components/schema/deps/docs reviewed; conflicts flagged).
  • Documents generated with correct template, full sections, and DRY warnings (if applicable).
  • Type validation passed (no test creation for impl; regression strategy for refactor; risk matrix/limits for test).
  • Preview shown with DRY warnings count and user confirmed.
  • Linear issues created with parentId and URLs captured; state=Backlog.
  • kanban_board.md updated under correct Epic/Story with indentation.
  • Summary returned with URLs, totals, DRY warnings count, and next steps.

Template Loading

MANDATORY READ: Load shared/references/template_loading_pattern.md for template copy workflow.

Template Selection by taskType:

  • implementationtask_template_implementation.md
  • refactoringrefactoring_task_template.md
  • testtest_task_template.md

Local copies: docs/templates/*.md (in target project)

Reference Files

  • Tools config: shared/references/tools_config_guide.md
  • Storage mode operations: shared/references/storage_mode_detection.md
  • Kanban update algorithm: shared/references/kanban_update_algorithm.md
  • Template loading: shared/references/template_loading_pattern.md
  • Linear creation workflow: shared/references/linear_creation_workflow.md
  • Templates (centralized): shared/templates/task_template_implementation.md, shared/templates/refactoring_task_template.md, shared/templates/test_task_template.md
  • Local copies: docs/templates/*.md (in target project)
  • Kanban format: docs/tasks/kanban_board.md

Version: 3.0.0 Last Updated: 2025-12-23

相关 Skills

PPT处理

by anthropics

Universal
热门

处理 .pptx 全流程:创建演示文稿、提取和解析幻灯片内容、批量修改现有文件,支持模板套用、合并拆分、备注评论与版式调整。

涉及PPTX的创建、解析、修改到合并拆分都能一站搞定,连备注、模板和评论也能处理,做演示文稿特别省心。

效率与工作流
未扫描119.1k

技能工坊

by anthropics

Universal
热门

覆盖 Skill 从创建到迭代优化全流程:起草能力、补测试提示、跑评测与基准方差分析,并持续改写内容和描述,提升效果与触发准确率。

技能工坊把技能从创建、迭代到评测串成闭环,方差分析加描述优化,特别适合把触发准确率打磨得更稳。

效率与工作流
未扫描119.1k

Word文档

by anthropics

Universal
热门

覆盖Word/.docx文档的创建、读取、编辑与重排,适合生成报告、备忘录、信函和模板,也能处理目录、页眉页脚、页码、图片替换、查找替换、修订批注及内容提取整理。

搞定 .docx 的创建、改写与精排版,目录、批量替换、批注修订和图片更新都能自动化,做正式文档尤其省心。

效率与工作流
未扫描119.1k

相关 MCP 服务

文件系统

编辑精选

by Anthropic

热门

Filesystem 是 MCP 官方参考服务器,让 LLM 安全读写本地文件系统。

这个服务器解决了让 Claude 直接操作本地文件的痛点,比如自动整理文档或生成代码文件。适合需要自动化文件处理的开发者,但注意它只是参考实现,生产环境需自行加固安全。

效率与工作流
83.9k

by wonderwhy-er

热门

Desktop Commander 是让 AI 直接执行终端命令、管理文件和进程的 MCP 服务器。

这工具解决了 AI 无法直接操作本地环境的痛点,适合需要自动化脚本调试或文件批量处理的开发者。它能让你用自然语言指挥终端,但权限控制需谨慎,毕竟让 AI 执行 rm -rf 可不是闹着玩的。

效率与工作流
5.9k

EdgarTools

编辑精选

by dgunning

热门

EdgarTools 是无需 API 密钥即可解析 SEC EDGAR 财报的开源 Python 库。

这个工具解决了金融数据获取的痛点——直接让 AI 读取结构化财报,比如让 Claude 分析苹果的 10-K 文件。适合量化分析师或金融开发者快速构建数据管道。但注意,它依赖 SEC 网站稳定性,高峰期可能延迟。

效率与工作流
2.0k

评论