Generate structured Task Prompts and coordinate P8 agent teams for multi-agent delivery.
npx skills add https://github.com/sandmark78/workspace --skill p9-sandmark78
Please help me install this Agent Skill. Skill: p9 Source: https://github.com/sandmark78/workspace/tree/main/skills/p9 Command: npx skills add https://github.com/sandmark78/workspace --skill p9-sandmark78
P9 mode acts as a high-level tech-lead assistant that writes Task Prompts and coordinates P8 agent teams without writing code itself, enabling scalable project orchestration through structured prompts.
Tell the AI to switch into P9 mode and begin coordinating a multi-agent project by providing a project brief and milestones.
High-intent search queries and answers about installing and using this skill.
You can coordinate multiple AI agents by using a tech-lead assistant to generate structured Task Prompts and oversee P8 teams. This method handles task decomposition, delegation, and delivery for 3+ parallel agents without writing code directly.
A Task Prompt is a structured prompt containing six elements (六要素) that define goals, inputs, outputs, constraints, steps, and roles. It manages agent teams by providing clear task decomposition and delegation parameters for P8-level execution.
You need prompt engineering for multi-agent project delivery when coordinating 3 or more parallel agents across a software project. It allows you to create an orchestration plan, assign tasks, track progress, and consolidate results via the P9 protocol.
To start leading a project with prompts, switch the AI into P9 mode and provide a project brief along with milestones. The system then generates an orchestration plan and relays structured Task Prompts to P8 teams for execution.
The P9 protocol is designed for scenarios with three or more parallel agents, covering task decomposition, delegation, and delivery. It is not intended for smaller teams, as its orchestration features specifically target complex multi-agent coordination.
cshyangPlan and execute a project discovery phase to define requirements and scope.
caobingshengCoordinate multiple AI agents to execute complex project tasks without coding.
handsomelong922Design an implementation plan, code results, and deliver a 3-question self-review via P7-COMPLETION.
CodingHeaderCoordinate LLM agents to work in parallel on independent software-engineering tasks, aggregating their results and monitoring progress along the way.
