task▌

ln-302-task-replanner

Structured task planning with user stories, sprint organization, and backlog prioritization for agile teams. \n \n Provides templates for writing INVEST-compliant user stories with acceptance criteria, technical notes, estimation, and dependencies \n Includes epic decomposition patterns breaking features into stories and granular tasks with time estimates \n Supports MoSCoW prioritization (Must/Should/Could/Won't Have) and sprint planning with capacity planning and Definition of Done checklists

CRITICAL — 开始前 MUST 先用 Read 工具读取 ../lark-shared/SKILL.md，其中包含认证、权限处理

Never Use TaskOutput \n TaskOutput floods the main context window with agent transcripts (70k+ tokens). \n Rule \n NEVER use TaskOutput tool. Use Task tool with synchronous mode instead. \n Why \n \n TaskOutput reads full agent transcript into context \n This causes mid-conversation compaction \n Defeats the purpose of agent context isolation \n \n Pattern \n # WRONG - floods context\n Task(run_in_background=true)\n TaskOutput(task_id=\"...\") // 70k tokens dumped\n \n # RIGHT - isolated context

Decompose complex tasks, design dependency graphs, and coordinate multi-agent work with structured task descriptions. \n \n Provides four decomposition strategies: by architectural layer, functional component, cross-cutting concern, or file ownership to parallelize work effectively \n Includes dependency graph patterns (independent, sequential, diamond) with principles for minimizing chain depth and identifying critical paths \n Offers task description template covering objective, owned files, r