Agents & System Prompts
301 results
- Objective Expert CollaboratorSystem prompt that removes flattery, holds positions unless given new evidence, admits unknowns and ends each reply with counterpoints.
- Opus-Driven Deep Thinking SystemFrames the model as a role-based system where Opus makes decisions, Sonnet handles development and Haiku does research, for deep thinking and building.
- Orchestration Agent (PowerPlatformSupervisor)A JSON system prompt for an orchestration agent that analyzes requests and routes each to the single best supervisor and sub-agent.
- Parallel Agents With GoalsTells the agent to split a task into independent pieces, spawn parallel sub-agents each with its own /goal, and merge their results.
- Performance MonitorAnalyzes metric, log and output files for performance patterns and anomalies, then writes an evidence-based observability plan in Markdown.
- Personal Finance AdvisorYou are a personal finance advisor, providing guidance on budgeting, saving, investing, and managing debt.
- Personal Growth CoachAs an AI Personal Growth Coach, your primary objective is to assist users in their journey of self-improvement and personal development
- Personalized GPT Assistant PromptShort system prompt for an assistant that adapts its tone, topic focus and default language to user preferences.
- Philosophical Analysis AssistantSpecializes in Kantian and Hegelian philosophical analysis consultations, fostering critical thinking
- Plan Check LoopMakes the agent re-examine its own strategy or plan, find loopholes, propose fixes and repeat until it is confident the plan holds.
- Plan Mode: Strategic Planning & ArchitectureAgent mode that analyzes the codebase and clarifies requirements before any code is written, then proposes an implementation strategy.
- Plan-and-Act Research AgentGuides a tool-using agent through a plan-act-reflect research loop, requiring every conclusion to trace back to evidence returned by tools.
- Planning ModeShort agent mode that writes an implementation plan in Markdown (overview, requirements, steps, tests) for a feature or refactor without editing code.
- Principled Problem SolverExcellent at principled problem-solving and categorization. Chain of Thought agent
- Productive Peer MentorA persona prompt for a friendly, tech-savvy peer mentor who shares productivity habits naturally without forcing tech into casual topics.
- Programming Team BuilderDefines a programming team of roles such as team brain, task distributor, programmers and manager, and how they collaborate.
- Prompt 101 (Full Template)A fill-in prompt skeleton covering role, context, goal, tone, reference documents, task rules, examples and conversation history.
- Prompt Architect ProTurns the model into a prompt engineer that asks clarifying questions, writes structured system prompts with a fixed framework and explains the techniques used.
- Prompt BuilderTwo-persona agent that drafts prompts and then tests them, analyzing requirements and refining structure until the prompt meets prompt-engineering best practices.
- Prompt CreatorAssume the role of my 'Prompt Engineer,' tasked with aiding me in designing an optimal, personalized prompt that suits my needs perfectly.
- Prompt EngineerSets the model up as a prompt engineer that turns a user goal into a structured, engineered prompt for a target AI model.
- Prompt EngineerAgent persona for designing, testing and evaluating LLM prompts in production, including A/B tests and token cost control.
- Prompt EngineerTreats every input as a prompt to improve: first analyzes it in a reasoning block against a fixed framework, then outputs a rewritten system prompt.
- Prompt EnhancerRewrites simple prompts into more detailed, thought-provoking questions and explains the enhancement process with an example.