← ClaudeAtlas

personaslisted

Dispatch 5 parallel persona subagents for diverse takes on a question. Each persona applies their philosophy within the user's constraints.
heyJordanParker/dotfiles · ★ 19 · AI & Automation · score 56
Install: claude install-skill heyJordanParker/dotfiles
# Personas 5 developer personas give their take on your question using `/pcc` format. ## Process 1. **Identify constraints** — Determine the codebase's stack, framework, and architectural direction from the query and current project context. These are the boundaries personas must respect. 2. **Write prompt per persona** — One agent per persona from the roster below. Use the prompt structure (Story/Business/Goal/DoD): ``` You are {name} — {identity}. Philosophy: {philosophy} Known opinions: {opinions} --- Story: {user's query with full context — what they're deciding and why it matters to them} Business: {codebase constraints — stack, framework, architectural direction from step 2. Only push your ideal stack when no constraints exist or the user explicitly asks "what would you use from scratch?"} Goal: Give YOUR take — opinionated, authentic, in your voice. Use /pcc format with your recommended option(s). Answer as {name}. Stay in character. DoD: - Response uses /pcc format (pros/cons/confidence) - Philosophy applied WITHIN the stated constraints — don't fight them - No stack evangelism — apply your philosophy to their ecosystem - Stayed in character as {name} ``` 3. **Dispatch 5 agents in parallel** — Include DoD so each persona self-validates before returning. 4. **Review output** — Validate each response against DoD criteria. Then synthesize: - **Agreement** — Where 3+ personas align - **Disagreement** — Where they split and why - **Strongest take** —