TryCrucible — Prove your AI engineering skills

AllRAG Pipelines AI Agents MCP Servers Coding Agents Evals & Testing AI Tool Proficiency

// 12 challenges · hard

Build a coding agent that fixes failing CI

Build a coding agent that reads failing CI logs, identifies the code defect, patches the repository, and adds or preserves tests. The target repositories are Python — your agent can be implemented in any language.

Build a coding agent that migrates Flask routes to FastAPI

Build a coding agent that migrates Flask API endpoints to FastAPI — converting route handlers to async, adding Pydantic models, and using dependency injection. The target codebase is Python — your agent can be implemented in any language.

Challenges

Build a coding agent that fixes failing CI

Build a coding agent that migrates Flask routes to FastAPI

Build a multi-agent research pipeline

Build an MCP server for a REST API with auth and rate limiting

Debug a performance incident using AI tools

Debug a production incident using AI tools only

Extract structured data from multimodal documents

Production RAG with hybrid search and re-ranking

Review and patch security vulnerabilities using AI tools

Write evals for a summarisation model

Write evals for prompt injection defense

Write evals for RAG answer faithfulness