All prompts

Design an AI Agent workflow that can automatically read ML papers, extract key implementation details, and generate runnable reproduction code and experiment configurations.

Use AI to quickly identify the true file type to prevent malicious files from disguising themselves, suitable for security reviews and data pipeline preprocessing scenarios.

Design an end-to-end AI Agent workflow that automatically reads ML papers, extracts experimental settings, reproduces the training process, and generates evaluation reports.

Let AI help you design a complete ML paper reproduction workflow: from paper interpretation to code implementation, data preparation, training tuning, and result comparison, suitable for researchers quickly validating cutting-edge papers.

Design an end-to-end agentic commerce workflow where an AI Agent autonomously completes product discovery, price comparison, and ordering, suitable for e-commerce automation and intelligent procurement scenarios.

Describe operator requirements in natural language to automatically generate TileLang DSL code, supporting rapid development of common high-performance computing kernels such as GEMM, Attention, and quantization.

Automatically scans the code warehouse for security risks, architectural issues, and technical debt, generating a prioritized repair roadmap and health score

Simulates business negotiation scenarios between two AI Agents to help understand automated business agreements and multi-party game strategies

Generates a unified compatibility layer design solution for any AI model API, supporting multi-provider routing and format conversion

Design a comprehensive observability solution for the AI Agent's multi-step task execution chain, including trace, span, log association, and performance metric collection

Using recursive language models (RLM) to design iterative execution schemes for complex reasoning tasks, including sandbox isolation and quality assessment

Generate standardized code review workflows for multi-agent coding teams, including role division, review rules, and quality access controls

Paste your prompt, and AI analyzes it from perspectives such as clarity, structure, constraints, and provides optimized versions and test cases

Design a secure execution scheme for Python code generated by AI Agents, including sandbox isolation, permission control, timeout mechanisms, and output validation, suitable for production scenarios requiring LLM code execution.

Input a GitHub repository URL to automatically generate a quick start guide for first-time contributors, including environment setup, code structure, and recommended directions for the first PR.

Generates natural language motion descriptions for AI-driven character animation systems, outputting structured motion sequence instructions usable for neural network animation models

Conduct quality reviews of entire LLM-translated books, checking terminology consistency, stylistic consistency, and cultural fit, and provide revision suggestions

Design a fully self-hosted, privacy-first personal knowledge management system integrating local AI for semantic search, automatic tagging, knowledge graphs, and intelligent summarization, with all data remaining local.

Design a locally deployed multi-model collaborative inference solution supporting models such as large and small model cascading, routing distribution, and result fusion, maximizing inference efficiency and quality balance

A comprehensive guide for AI short film production, from scriptwriting and storyboarding to image generation, voiceover, and music, applicable to video generation tools like Sora, Kling, and others.

Design an architecture to unify batch processing, stream processing, and AI inference workloads into a single high-performance compute engine, replacing traditional distributed deployments of Spark and Flink.

Automated PR lifecycle management from idea to merge, including task assignment, code generation, automatic review, conflict resolution, and merge strategies.

Input a theme and mood to generate lyrics and style tag descriptions suitable for AI music generation tools (such as ACE-Step, Suno).

Convert meeting audio/text records into structured meeting minutes, automatically extracting action items, decisions, and follow-up tasks.

Automatically generate standardized Pull Request descriptions based on code changes, including change summaries, impact analysis, and testing suggestions.

Automatically rewrite a core piece of content into copy adapted for different platforms such as Xiaohongshu, Twitter, LinkedIn, and WeChat Official Accounts.

Quickly layout your research content into a professional academic paper format using Markdown, supporting citations, formulas, figure/table numbering, and automatic reference generation.

Automatically generates content variants, publishing schedules, and engagement strategies adapted for platforms like Xiaohongshu, Twitter/X, LinkedIn, and WeChat Official Accounts based on a core content theme.

Design an AI-based automatic issue triage system for open source project maintainers, including classification rules, priority assessment matrices, and auto-reply templates.

Evaluate the feasibility of running Windows applications on Linux and generate detailed migration plans, including containerization, Wine, native alternatives, and other path comparisons and recommendations. Suitable for teams considering migrating from Windows to Linux.

Converts Git diffs or code changes into clear natural language explanations, helping reviewers quickly understand intent and scope.

Uses LLM multi-agents for automated quantitative factor mining: a complete workflow design prompt covering data cleaning, factor construction, IC analysis, and strategy backtesting, adaptable to platforms like Qlib/RD-Agent.

Before committing code, let AI analyze the modules, dependency chains, and potential risk points affected by the changes to avoid deployment incidents.

Design a local knowledge base management solution based on Markdown files, combined with AI to achieve intelligent retrieval, automatic tagging, bidirectional link recommendations, and knowledge graph visualization

Cleans chat records from WeChat/QQ into high-quality fine-tuned datasets to train AI digital avatars with personal styles. Automatically filters noise, extracts dialogue patterns, and generates instruction-tuning format data.

Design an AI Agent system capable of autonomous learning and skill accumulation, allowing the Agent to automatically build a skill library through task execution.

Upload suspicious file descriptions or characteristics; AI analyzes the true file type, detects spoofing, assesses potential security risks, and provides disposal recommendations.

Converts natural language descriptions into precise Shell commands, automatically performing security risk assessment and parameter validation to avoid dangerous operations.

Design an AI Agent self-evolution framework based on Gene Expression Programming (GEP), allowing the Agent to autonomously grow its skill tree.

Analyze your prompt templates to identify logic flaws, ambiguous expressions, and token waste, then output an optimized version.

Design a foundation model fine-tuning plan for financial time-series data, including data preprocessing, training strategies, and backtesting frameworks.

Describe your application interface and user flows; AI generates end-to-end automated test scripts based on visual recognition, without relying on DOM selectors.

Input multiple competitor names to automatically generate a detailed feature comparison matrix table, including pricing, core features, tech stack, and pros/cons analysis.

Design anomaly detection rules for AI Agent systems, including monitoring strategies for abnormal token consumption, tool call anomalies, loop detection, and security boundary breaches.

Input a one-sentence description of the MCP Server functionality you want, and automatically generate a complete design document, including tool definitions, resource schemas, transport protocol selection, and implementation code framework.

A complete self-evolution loop prompt that enables AI Agents to autonomously discover new skill needs, generate skill code, and validate effects.

Input image/video/document descriptions to automatically generate multi-dimensional structured tags, summaries, and knowledge graph nodes.

Generate task assignment rules, file locking strategies, and code merge conflict prevention plans for multi-agent coding teams, enabling AI coding teams to collaborate efficiently like real development teams.