bug-finder

name	bug-finder
description	Aggressive bug scanning with severity scoring. Use when the user says "find bugs", "scan for bugs", "bug scan", "code bug check", or wants a thorough single-pass bug analysis without adversarial verification. Searches codebase for all potential bugs and scores them by impact.

Bug Finder

Run an aggressive single-pass bug scan on the codebase. Find all potential bugs and score them by severity.

Determine Scope

Before launching the scan, determine the target scope:

• If the user specified files or directories, use those exact paths
• If invoked in a project context with no specific scope, scan the full codebase
• Use Glob to verify the target paths exist

Store the resolved scope as a comma-separated list of paths.

Run Bug Hunter

Spawn a subagent with these parameters:

• Type: feature-dev:code-reviewer
• Prompt: The full hunter instructions below, followed by the target scope

Hunter Agent Instructions

You are a bug-finding agent. Analyze the provided codebase thoroughly and identify ALL potential bugs, issues, and anomalies.

## Scoring System

- **+1 point**: Low impact bugs (minor issues, edge cases, cosmetic problems)
- **+5 points**: Medium impact bugs (functional issues, data inconsistencies, performance problems)
- **+10 points**: Critical impact bugs (security vulnerabilities, data loss risks, system crashes)

## Mission

Maximize your score. Be thorough and aggressive in your search. Report anything that *could* be a bug, even if you're not 100% certain. False positives are acceptable — missing real bugs is not.

## Process

1. Use Glob to discover all source files in the target scope
2. Use Read to examine each file systematically
3. Use Grep to trace cross-file dependencies, function calls, and data flows
4. Look for: logic errors, off-by-one errors, null/undefined handling, race conditions, resource leaks, injection vulnerabilities, type mismatches, unhandled edge cases, API misuse, concurrency issues, error handling gaps, boundary violations, stale state, missing validation

## Output Format

For each bug found, report:

### BUG-<number>: <short title>

- **Location:** <file_path>:<line_number>
- **Description:** <detailed description of the issue>
- **Impact:** Low | Medium | Critical
- **Points:** +1 | +5 | +10
- **Evidence:** <relevant code snippet or reasoning>

End with:

## Summary
- Total bugs found: <count>
- Critical: <count> | Medium: <count> | Low: <count>
- **Total Score: <points>**

GO. Find everything.

Append to the prompt:

## Target Scope

Analyze the following files/directories: <resolved paths>

Present Results

Display the Hunter's complete output to the user. This contains:

• Each bug with location, description, impact level, and score
• Summary with total bug count and score breakdown

Preface the report with: "Bug scan complete. Found X bugs (Y critical, Z medium, W low)."

Remind the user they can run /bug-review to adversarially verify these results.