agent-prompt-design
Create well-structured prompts for AI agents using proven architecture patterns. Use when users ask to write agent prompts, system prompts, or agent instructions, or want to improve existing prompts that aren't working.
npx skills add majiayu000/claude-skill-registry --skill agent-prompt-design --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# Agent Prompt Design ## Process ### Phase 1: Gather Requirements Collect the following before drafting: **Required:** - Agent's purpose and role (what it does) - Available tools and their capabilities - Primary tasks the agent should perform **If improving an existing prompt:** - Current prompt text - Specific failures or problems observed - Example queries that don't work well ### Phase 2: Draft the Prompt Structure Build the prompt with these five components in order: #### 2.1 Role Definition Write 1-2 sentences establishing identity and function. ``` You are a [role] that [primary function]. Your goal is to [main objective]. ``` #### 2.2 Dynamic Content Section Add placeholders for context that will be injected at runtime. Focus on domain-specific data the framework won't provide automatically: ``` ## Current Context - User: {{user_name}} - Account type: {{account_tier}} - Permissions: {{user_permissions}} ``` Note: Conversation history and tool definitions are typically handled by the framework—don't include them unless the system requires manual injection. #### 2.3 Detailed Instructions Write step-by-step behavioral guidance. Be specific about: - What to do first when receivi
- Process
- Phase 1: Gather Requirements
- Phase 2: Draft the Prompt Structure
- Phase 3: Add Explicit Heuristics
- Phase 4: Handle Tool Usage
- Phase 5: Add Safety Guardrails
- Phase 6: Validate the Prompt
- Failure Modes to Avoid
What does the agent-prompt-design skill do?
Create well-structured prompts for AI agents using proven architecture patterns. Use when users ask to write agent prompts, system prompts, or agent instructions, or want to improve existing prompts that aren't working.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill agent-prompt-design --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.