| name | advanced-prompt-engineering-suite |
| description | Provides procedural implementation and templates for creating advanced prompts, incorporating psychological principles and thinking methodologies. |
Advanced Prompt Engineering Suite (Skill)
Core Objective
To provide the procedural implementation and templates for creating and applying advanced prompts, incorporating psychological principles and specific thinking methodologies, as guided by the Prompt Engineering Specialist Agent.
Design Rationale (CoT/ToT Inspired):
- Problem: Crafting effective prompts requires specific syntax, structuring, and iterative refinement. Simply knowing what prompt to use is less helpful than knowing how to construct it reliably.
- Need: This skill encapsulates the concrete steps, templates, and patterns for prompt engineering, making it a reusable capability.
- Foundation: It operationalizes the design principles and strategies developed by the Prompt Engineering Specialist Agent, allowing other agents and skills to easily implement sophisticated prompting.
- Enables: Consistent and effective LLM interactions across the system, directly supporting the "prompt design" and "thinking methodology" requirements.
Key Functionalities:
- Generate prompt structures based on task type, desired LLM model, and cognitive goals.
- Incorporate user-defined context, examples, and constraints into prompts.
- Implement specific prompt patterns (e.g., Few-shot learning, role-playing, Chain-of-Thought structures).
- Provide templates for persona-based prompting or style emulation.
- Guide the formatting of inputs for multimodal models if applicable.
- Assist in formatting prompts for specific MCPs or API endpoints.
Tool & MCP Integration:
mcp_openrouter_chat_with_model: To generate prompt variations and test them.write_file / replace: To create prompt template files or incorporate prompt snippets into agent/skill code.save_memory: To store prompt templates and successful prompt strategies.
Dependencies & Enablers:
- Depends on: Guidance and strategies from the Advanced Prompt Engineering & Psychology Specialist Agent.
- Enables: Any agent or skill that needs to interact with an LLM in a sophisticated and controlled manner.
Blueprint Notes:
- This skill could be modular, with sub-skills for different prompt types (e.g.,
prompt-suite.cot.skill, prompt-suite.persona.skill).
- It should provide clear examples and usage instructions.
- Its effectiveness will be reviewed by the Quality Assurance & Auditing Bureau Agent.
