Overview

The Coding Assistant helps you write, review, and debug code across languages. It emphasizes understanding the request, inspecting existing code with read_file before editing, making small focused changes, and explaining what changed and why. When reviewing or debugging, it highlights bugs, security issues, and performance problems with concrete fixes.

How This Skill Works

On a request, it first asks clarifying questions to resolve ambiguity. It then uses read_file to understand the current code before making edits, applying small, focused changes and avoiding unrelated refactors. It explains each change, suggests tests, and, during debugging, asks for error messages, uses execute_command to run logs or tests, and presents the root cause along with a fix.

When to Use It

• You’re adding a new function or feature and want clean, minimal changes.
• You need a code review that spots bugs, security gaps, and performance issues.
• You’re debugging and can provide error messages and reproduction steps.
• You want small, localized fixes rather than broad refactors.
• You want a clear explanation of what changed, why, and how to test it.

Quick Start

1. Step 1: Clarify the request and confirm goals.
2. Step 2: Read the existing code with read_file, then draft a focused change.
3. Step 3: Implement the change, explain what and why, and provide test steps.

Best Practices

• Ask clarifying questions before touching code to avoid scope creep.
• Read the existing code with read_file before proposing edits.
• Make small, focused changes and avoid refactoring unrelated parts.
• Explain each change and the rationale behind it.
• Provide concrete tests or test steps and how to run them.

Example Use Cases

• Add input validation and type checks to a function with a minimal, localized patch.
• Review a module to identify a security flaw and patch it with a fixed code snippet and inline comments.
• Identify a performance bottleneck in a loop and propose a small optimization with before/after timing.
• Debug a failing test by requesting error messages and reproduction steps, then fix and re-run tests.
• Refactor a small utility function with a focused change and explain the impact on downstream code.

Frequently Asked Questions