Generate Optimized C++ Competitive Programming Solutions

This agent operates as an autonomous competitive programming specialist, analyzing algorithmic problems and generating optimized C++ solutions with explanations and complexity analysis. It runs a five-stage process: problem analysis (parsing input/output format, constraints, and edge cases, categorizing the problem type - graph, DP, greedy, math, string - and identifying target time/space complexity from the constraints), algorithm design (generating multiple approaches - brute force, optimized, alternative methods - and selecting the most efficient one given time and memory limits), implementation (writing clean, optimized C++ with fast I/O, appropriate data types, and inline comments on complex logic), verification (manually tracing provided examples, checking edge cases like empty input or maximum constraints, and validating complexity against problem limits), and documentation (explaining the approach in clear terms with best/average/worst-case complexity analysis and any relevant alternative approaches).

It outputs a complete, runnable C++ solution using a fast I/O setup, an algorithm explanation with time and space complexity, and alternative approaches when applicable.

int main() {
    ios_base::sync_with_stdio(false);
    cin.tie(NULL);
    
    return 0;
}

Use this agent when you need a competitive-programming-style C++ solution to an algorithmic problem - one intended to pass judge systems like Codeforces, AtCoder, or LeetCode - with complexity analysis and edge-case verification included.

It is not a fit for production application code, where the guidelines' priorities (terse-but-fast STL usage, minimal abstraction, judge-oriented I/O) trade off against the maintainability and API design concerns production code usually needs.

Inputs are the problem statement, including its constraints and expected input/output format. Outputs are a complete C++ solution with fast I/O, an algorithm explanation (approach, key insights, time/space complexity), and alternative approaches when more than one viable method exists.

Solutions are written to pass judge systems such as Codeforces, AtCoder, or LeetCode.

Competitive programmers and anyone practicing algorithmic problem-solving who wants a complete, judge-ready C++ solution with explicit complexity analysis - covering data type pitfalls (long long for large numbers, integer overflow), STL usage, and both adjacency-list and matrix representations for graph problems - rather than writing and verifying the solution from scratch.