Code Fix Assistant Skill

Code Fix Assistant

Automated code quality assistant providing a complete fix pipeline: formatting → syntax/type fixing → bug detection → validation, supporting multiple programming languages.

Capabilities

Code Formatting: Automatically apply standard code style and formatting rules
- Python: Black, autopep8, isort
- JavaScript/TypeScript: Prettier, ESLint
- Java: Google Java Format
- Go: gofmt, goimports
- Rust: rustfmt
Syntax/Type Error Fixing: Detect and fix common syntax and type issues
- Missing semicolons, unmatched brackets
- Import/dependency issues
- Type errors (TypeScript, Java, Rust)
- Unused variables and imports
Bug Detection & Fixing: Static analysis and intelligent fixes
- Null pointer/null reference issues
- Array out of bounds
- Resource leaks (files, connections)
- Logic errors (incorrect conditions)
- Dead code detection
Fix Validation: Automatically validate fix effectiveness
- Run linter checks
- Syntax validation
- Generate fix reports
- Compare before/after differences

Input Requirements

Required Inputs:

file_path: Path to code file to fix
language: Programming language (python, javascript, typescript, java, go, rust)
fix_types: List of fix types (optional, defaults to all)
- format: Code formatting
- syntax: Syntax/type fixing
- bugs: Bug detection and fixing
- validate: Fix validation

Optional Inputs:

config: Custom configuration (e.g., ESLint rules, Black options)
severity_threshold: Severity threshold (error, warning, info)
auto_apply: Whether to auto-apply fixes (default false, report only)

Input Format:

{
  "file_path": "src/main.py",
  "language": "python",
  "fix_types": ["format", "syntax", "bugs", "validate"],
  "auto_apply": true,
  "severity_threshold": "warning"
}

Output Formats

Fix Report (JSON format):

{
  "file": "src/main.py",
  "language": "python",
  "fixes_applied": {
    "format": { "changes": 15, "status": "success" },
    "syntax": { "issues_found": 3, "issues_fixed": 3 },
    "bugs": { "issues_found": 2, "issues_fixed": 1, "unfixable": 1 },
    "validation": { "linter_passed": true, "syntax_valid": true }
  },
  "issues": [
    {
      "type": "syntax",
      "severity": "error",
      "line": 42,
      "message": "Undefined variable 'results'",
      "fix": "Added variable initialization",
      "status": "fixed"
    },
    {
      "type": "bug",
      "severity": "warning",
      "line": 58,
      "message": "Potential null pointer dereference",
      "fix": "Added null check",
      "status": "fixed"
    }
  ],
  "diff": "--- original\n+++ fixed\n@@ -42,3 +42,4 @@\n+results = []\n",
  "summary": "Applied 15 formatting changes, fixed 3 syntax errors, fixed 1/2 bugs, validation passed"
}

Fixed Code File (if auto_apply: true)

How to Use

Basic Usage:

@code-fix-assistant

Fix formatting and syntax issues in this Python file: src/main.py

Complete Fix Pipeline:

@code-fix-assistant

Run complete fix pipeline on src/app.ts:
1. Format code
2. Fix TypeScript type errors
3. Detect and fix potential bugs
4. Validate fix results

Batch Fixing:

@code-fix-assistant

Fix code quality issues in all Python files under src/ directory

Check Only (No Auto-Fix):

@code-fix-assistant

Check code issues in main.go and generate report, but don't auto-fix

Scripts

code_fixer.py: Main fix engine coordinating all fix steps
formatters.py: Code formatting module (multi-language support)
syntax_checker.py: Syntax and type error detection & fixing
bug_detector.py: Static analysis and bug detection
validator.py: Fix validation and report generation
language_config.py: Language-specific config and tool mapping

Best Practices

Incremental Fixing: Recommended order: format → syntax → bugs
Version Control: Commit code before fixing for easy rollback
Test Validation: Run test suites after fixing to ensure no functionality breakage
Review Fixes: For critical code, generate report and manually review before applying
Config Files: Use project-level configs (.prettierrc, .black.toml) for consistency
Batch Fixing: For large projects, fix incrementally to avoid massive changesets

Limitations

Tool Dependencies:

Requires language-specific linter/formatter tools installed
Python: black, flake8, mypy
JavaScript/TypeScript: prettier, eslint
Java: google-java-format, checkstyle
Go: gofmt, golangci-lint
Rust: rustfmt, clippy

Fix Capabilities:

Can only fix automatable issues (formatting, common syntax errors, simple bugs)
Complex logic errors require human intervention
Cannot guarantee 100% fix rate for all issues
Some bugs require contextual understanding and may be misdiagnosed

Performance Considerations:

Large files (>10,000 lines) may be slow to process
Batch fixing many files takes time
Static analysis depth is limited (not full program analysis)

Security:

Will not modify .git/ or config files
Will not execute code (static analysis only)
Recommended to test in isolated environment
Critical production code should be manually reviewed

Language-Specific Notes

Python

Default uses Black format (88-character line width)
Supports type hint checking (mypy)
Auto-sorts imports (isort)

JavaScript/TypeScript

Follows Prettier default rules
ESLint rules customizable
TypeScript type errors auto-fixed (simple types only)

Java

Google Java Format style
Supports null checking and resource leak detection
Unused imports auto-cleaned

Go

Strictly follows gofmt standard
goimports auto-manages imports
Supports go vet static checks

Rust

rustfmt standard formatting
Clippy lint suggestion fixes
Ownership and lifetime errors reported (not auto-fixable)

Error Handling

Tools Not Installed:

Prompts to install missing tools
Provides installation commands
Skips step and continues with other fixes

Fix Failures:

Logs failure reason
Preserves original file unchanged
Provides manual fix suggestions

File Permission Issues:

Checks read/write permissions
Prompts permission errors
Suggests solutions

Examples

Example 1: Python Formatting and Syntax Fixing

Input:

# bad_code.py
import os,sys
import json

def process_data(data):
  results=[]
  for item in data:
      if item['value']>0:
        results.append(item)
  return results

def main( ):
    data=[{'value':1},{'value':-1}]
    result=process_data(data)
    print(result)

After Fix:

# bad_code.py
import json
import os
import sys


def process_data(data):
    results = []
    for item in data:
        if item["value"] > 0:
            results.append(item)
    return results


def main():
    data = [{"value": 1}, {"value": -1}]
    result = process_data(data)
    print(result)

Example 2: TypeScript Type Error Fixing

Input:

// app.ts
function greet(name) {
  return "Hello, " + name;
}

let user = { name: "Alice", age: 30 };
console.log(greet(user.name));
console.log(user.address); // Error: Property 'address' does not exist

After Fix:

// app.ts
function greet(name: string): string {
  return "Hello, " + name;
}

interface User {
  name: string;
  age: number;
}

let user: User = { name: "Alice", age: 30 };
console.log(greet(user.name));
// Removed: console.log(user.address); - Property does not exist

Example 3: Go Code Formatting and Lint Fixing

Input:

// main.go
package main
import "fmt"
func main(){
var x int=10
if x>5{
fmt.Println("x is greater than 5")
}
}

After Fix:

// main.go
package main

import "fmt"

func main() {
	var x int = 10
	if x > 5 {
		fmt.Println("x is greater than 5")
	}
}

Agent Skills: Code Fix Assistant

Install this agent skill to your local

Skill Files