All prompts

Design a multi-agent self-learning collaboration system based on the Swarm architecture, including task routing, agent communication protocols, memory sharing, and self-evolution mechanisms

Describe your task requirements, and the AI will analyze which model to use (GPT-4o/Claude/Gemini/DeepSeek/local models), offering recommendations and reasons from four dimensions: cost, speed, quality, and privacy.

Like OpenAI Symphony, project work is transformed into independent, autonomous tasks, allowing teams to manage work rather than supervise AI. Suitable for engineering teams that need multi-agent parallel development and automated CI/PR processes.

Guide to deploying efficient TTS speech synthesis services in a pure CPU environment, without the need for a GPU. Suitable for voice application development in edge devices, embedded systems, or cost-sensitive scenarios.

Based on project type and team specifications, automatically generates adaptive workflow lifecycle rule files adapted to AI coding agents (Claude Code, Codex, Cursor, etc.), including behavioral constraints for planning, implementation, testing, and review.

Evaluate whether a device is suitable for running local large models and generate a complete deployment plan including performance estimates and optimization suggestions.

Input exported WeChat chat data to automatically generate an annual social analysis report, including visual insights on chat frequency, sentiment trends, and key topics.

Guides you step-by-step through your first open-source contribution, from finding a project to submitting a PR.

Design a knowledge graph-driven intelligent Q&A system based on the GraphRAG architecture, supporting multi-hop reasoning and entity-relation extraction.

Generate fine-tuning plans for financial time-series foundation models, including data preparation, training strategies, and evaluation metrics.

Input a product name and a list of competitors to automatically generate a structured SWOT analysis report, featuring in-depth comparisons across strengths, weaknesses, opportunities, and threats.

Input a core message to instantly generate different styles of copy adapted for platforms like Xiaohongshu, Twitter, LinkedIn, and WeChat Official Accounts.

Systematically review code change plans generated by AI coding agents, providing structured feedback to ensure code quality and architectural consistency.

Transform vague feature requirements into precise technical specification documents, optimized for direct execution by AI coding assistants.

Design a complete workflow for task assignment, skill reuse, and progress tracking for AI coding agent teams, suitable for team-level agent management scenarios.

A complete workflow prompt for fine-tuning financial time series foundation models, feature engineering, and generating trading strategies.

System prompt template for centrally managing multiple AI coding agents, enabling automatic task assignment, progress tracking, and result aggregation.

Simulate a senior architect conducting a comprehensive review of technical solutions, providing scores and improvement suggestions from dimensions such as feasibility, scalability, security, and cost.

Input a note topic to automatically generate 10 Xiaohongshu-style viral titles and cover copies, including high click-through rate elements like emojis, numbers, and pain-point keywords.

Generates a weekly multi-platform content calendar including titles, copy, and optimal posting times based on topics and target audiences.

Analyzes technical debt in codebases, categorizes it by priority, and generates an executable repayment plan suitable for team sprint planning.

Simulate structured discussions among multiple expert Agent roles, analyzing issues from diverse perspectives to generate consensus solutions.

Quickly evaluate whether a GitHub open-source project is worth watching, outputting a structured score and a one-sentence conclusion.

Help you design an AI Agent capable of self-learning and growth, automatically expanding its skill tree starting from seed skills.

Automatically decompose tasks based on Product Requirements Documents (PRD) and assign them to multiple AI coding Agents for collaborative completion.

Help you optimize local LLM inference performance on Mac, including memory management, batch processing configuration, and SSD caching strategies.

Use AI to analyze prediction market events, generating probability assessments and trading strategy recommendations.

Help you design, develop, and debug Claude Code plugins, including complete project structures and best practices.

Deeply compare the performance, ecosystem, and applicable scenarios of Rust async runtimes like Tokio, async-std, and smol, outputting a selection report.

Describe your application requirements in natural language to automatically generate production-grade Dockerfiles, supporting multi-stage builds and security hardening.

Generate a complete automated PR review process for open-source projects, including code quality checks, security scanning, test coverage requirements, and auto-merge strategy configuration.

Input any content (text, image descriptions, video scripts) to automatically generate a multi-dimensional tag system, suitable for content management, recommendation systems, and SEO optimization.

Design a complete canary release strategy for LLM applications, including model version switching, traffic allocation strategies, rollback mechanisms, and effect evaluation metric systems.

A prompt template to help AI Agents automatically discover, evaluate, and learn new skills, intelligently recommending and generating skill extension plans based on current task requirements.

Describe the 3D scene you want in natural language, and the AI will help you generate detailed scene descriptions and parameter configurations usable for generative 3D models.

Design a graph-structured RAG system that combines knowledge graph entity-relation reasoning with vector retrieval to improve complex Q&A quality.

Automatically generate production-grade Docker Compose deployment files based on the open-source project's tech stack and dependencies, including environment variables, persistence, and health checks.

Translate English READMEs into Chinese (or reverse), retain the original format, badges, and links, and automatically adapt to localized expression habits to generate complete files that can be directly submitted for PR.

Input a project requirement description to automatically generate a complete GitHub Actions workflow YAML file, supporting the entire process of testing, building, and deploying.

Evaluate and design speculative decoding solutions for LLM inference, comparing the performance of different draft models and verification strategies.

Input multiple open-source TTS model names to automatically generate a multi-dimensional comparative evaluation framework covering audio quality, multilingual support, voice cloning, latency, and deployment difficulty.

Input AI application requirements to generate a deployment solution comparison covering Web/Desktop/Mobile, including a decision matrix for cost, performance, and privacy dimensions.

Input business scenarios or tech stacks to automatically generate Wardley Map strategic analysis, identifying component evolution stages, strategic opportunities, and risk points.

Configure Speculative Decoding for your LLM inference service to achieve 2-5x inference acceleration via methods like Block Diffusion, including model selection, parameter tuning, and performance benchmarking.

Input a topic to automatically generate a 5-10 minute two-host podcast script, including intro, core discussion, audience interaction segment, and outro, suitable for direct audio generation via TTS tools.

Input your LLM inference scenario parameters to generate a speculative decoding acceleration plan evaluation report, including draft model selection, batching strategy, expected speedup ratio, and deployment recommendations.

Input GitHub repository information to automatically generate production-grade docker-compose.yml, Dockerfile, .env templates, and health check configurations, supporting GPU passthrough and reverse proxies.

Assess security risks in AI agent sandbox environments and generate escape attack surface analysis and defense hardening plans.