Agents & System Prompts
301 results
- Caveman ModeA response style that cuts tokens with terse bullets, short sentences and no filler while keeping code output readable; use when you want brief answers.
- Caveman ModeMakes a coding agent answer in terse, low-token replies with minimal wording while keeping full tool access. Useful when you want brief output.
- ChatGPT Prompt GeneratorTurns a topic you send into an expanded role prompt that begins with "I want you to act as", guessing what you want to do with it.
- CodexerA Python research assistant persona that uses Context7 MCP tools and web search to look up library docs before writing code.
- Compliance AuditorAgent persona for regulatory compliance work across GDPR, CCPA, HIPAA, PCI DSS, SOC 2 and ISO, covering controls, evidence collection and audit prep.
- Comprehensive Research Prompt GeneratorExpands a simple topic into a structured deep-research prompt with inquiry categories, exclusion criteria and formatting requirements.
- Constraint Solver ModeMakes the model frame a problem as variables, constraints and objectives, compare feasible solutions and trade-offs, and state an optimal choice with confidence.
- Context ManagerOrganizes the file-based shared context of a multi-agent workflow: directory layout, naming conventions and how agents read and update state.
- Context Preservation and MigrationAsks the AI to produce a structured context artifact capturing progress, decisions and project structure so another session, platform or agent can continue the work.
- Context-Aware Email AssistantAn email assistant that reads context across browser tabs, sets a goal with the user first and asks for confirmation when under 95 percent certain.
- Continue and Recap AssistantSystem prompt for resuming earlier work: recap past actions and decisions, confirm the last known state, ask for missing details, then continue.
- Continuous Execution ModeInstructs an agent to keep working without stopping: pick the next most valuable action, apply fixes and improvements as it goes, and only stop when told to.
- Conversation Drifting? Pull It Back with a Mid-Chat System DirectiveIn long chats the AI often "forgets" your original ask or drifts in style. Instead of starting over, send a single correction directive: restate the current…
- Conversational AISets up a friendly, polite conversational assistant with configurable topic, language and tone.
- Copilot Process Tracking InstructionsInstructions that make Copilot follow strict, ordered phases and record its process in an editable tracking file, useful for reviewing or resuming work later.
- Cosmic SeerInterdimensional wisdom oracle, insight into the essence of life
- Cosplay-Uchiha MadaraContributed by @FOX.
- COSTAR Framework Prompt WriterExpert in creating prompts based on the COSTAR Framework
- Critical Thinking (DeepThink)A system prompt that makes the model deconstruct its first answer, audit it for bias and fallacies, and rebuild a final verdict on a topic.
- Critical Thinking ModePuts the agent in a questioning mode that keeps asking why, challenges assumptions and helps an engineer think through an approach without editing code.
- Custom Agent FoundryHelps design and write VS Code custom agents by gathering the role, tools and workflow requirements and producing a suitable agent configuration.
- Cyber Security SpecialistYou are a cyber security specialist, providing guidance on securing digital systems, networks, and data.
- Daily DoctorExpertise in surgical diagnosis and personalized health management
- Devil's AdvocateStress-tests your idea by raising one objection at a time, pointing out flaws, risks and edge cases until you have a stronger proposal.