Decompose complex tasks into agent-assigned subtasks with explicit skills.
npx skills add https://github.com/HahyeonJeon/gobbi --skill plan-hahyeonjeon
Please help me install this Agent Skill. Skill: _plan Source: https://github.com/HahyeonJeon/gobbi/tree/main/plugins/gobbi/skills/_plan Command: npx skills add https://github.com/HahyeonJeon/gobbi --skill plan-hahyeonjeon
This Skill enables structured planning by decomposing complex tasks into small, agent-assigned subtasks, ensuring explicit agent and skills assignments and providing a disciplined path from ideation to delegation.
Enter plan mode to start planning, explore the codebase, decompose tasks, and present the plan for user approval.
High-intent search queries and answers about installing and using this skill.
Task decomposition breaks complex codebase work into agent-assigned subtasks with explicit skills. You enter plan mode, explore the project, decompose tasks, and present the structured plan for approval before delegation.
Plan mode is a structured exploration phase for codebase planning. It allows you to research domain knowledge, decompose work into small tasks, and assign specific subagents and skills before implementation begins.
To prevent task misrouting, each decomposed subtask must explicitly specify the assigned subagent and required skills. This structured delegation enforces constraints to avoid overlaps during execution.
Yes, when domain knowledge is insufficient, you can spawn research agents before planning. This ensures you have the necessary context to decompose tasks and structure execution order accurately.
The best way to manage execution order is using structured plan mode. It enforces constraints to avoid overlaps and misrouting by explicitly assigning agents and skills to each decomposed subtask before collection planning.
oxgeneralCoordinate teams of AI agents for parallel codebase tasks.
LamseyahEliasCoordinate subagent-driven development with two-stage spec and quality reviews.
Mr-Fang-VLSIEnforce GT3 backside routing policy for CTS/PDN layer usage and backside-net checks.
tigredonorteApply small targeted changes on an existing branch without full workflow orchestration.
