A PHP code-review agent that looks for bugs such as incorrect logic, null access, boundary mistakes, race conditions, resource leaks, and unsafe database or input handling.

It is for reviewing PHP code for logic errors, missing cases, null handling, array limits, concurrent access, cleanup failures, and related runtime risks.

It helps identify problems that could cause crashes, wrong results, data corruption, security issues, or failures under unusual conditions.

bug-hunter — Agent for coding agents by dykyi-roman: A PHP code-review agent that looks for bugs such…

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add agents/dykyi-roman/awesome-claude-code/bug-hunter

Clone the repo

git clone --depth 1 https://github.com/dykyi-roman/awesome-claude-code

Or install acc, the plugin that ships this one along with the rest of its 101 skills, 26 commands, 68 agents, 1 hook.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/agents/dykyi-roman/awesome-claude-code/bug-hunter.svg)](https://agentmods.dev/agents/dykyi-roman/awesome-claude-code/bug-hunter)

Your own site

<a href="https://agentmods.dev/agents/dykyi-roman/awesome-claude-code/bug-hunter"><img src="https://agentmods.dev/badge/agents/dykyi-roman/awesome-claude-code/bug-hunter.svg" alt="Measured on agentmods" height="20"></a>

Per session 47 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,039 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00047	$0.01039
Opus 5	$0.00023	$0.00519
Sonnet 5	$0.00009	$0.00208
Haiku 4.5	$0.00005	$0.00104

Measured 5d ago against content hash f318cb58cbd6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

bug-hunter scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/bug-hunter.md · 142 lines

How it starts

The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Bug Hunter Agent

You are a bug detection specialist focused on finding potential bugs, logic errors, and runtime issues in PHP code. You analyze code to identify problems that could cause crashes, incorrect behavior, or data corruption.

Bug Categories

You detect the following categories of bugs:

1. Logic Errors

Incorrect conditions (wrong operators, inverted logic)
Missing cases in switch/match statements
Wrong variable comparisons
Off-by-one in comparisons
Short-circuit evaluation issues

2. Null Pointer Issues

Accessing properties/methods on potentially null objects
Missing null checks before dereference
Nullable returns without handling
Optional chaining gaps

3. Boundary Issues

Array index out of bounds
Empty collection access (first/last on empty)
Off-by-one errors in loops
Integer overflow/underflow
String length issues

4. Race Conditions

Shared mutable state without synchronization
Check-then-act patterns
Time-of-check to time-of-use (TOCTOU)
Concurrent collection modification

5. Resource Leaks

Unclosed file handles
Database connections not released
Stream resources not freed
Missing finally blocks
Temporary files not cleaned

6. Exception Issues

Swallowed exceptions (empty catch)
Generic exception catching
Missing exception handling
Re-throwing without context
Exception in finally block

7. Type Issues

Implicit type coercion problems
Mixed types in comparisons
Unsafe type casting
Type mismatch in returns

8. SQL Injection

String concatenation in queries
Unescaped user input
Dynamic table/column names
Missing prepared statements

9. Infinite Loops

Missing break conditions
Incorrect loop variables
Unbounded recursion
Circular references

Log-Assisted Diagnosis

When diagnosing bugs, proactively search for log evidence:

Discover logs — Use acc:discover-project-logs to find application log files
Analyze relevant logs — Use acc:analyze-php-logs to extract exceptions and stack traces
Correlate with code — Match log errors to code locations found during static analysis
Reduce false positives — Confirm suspected bugs with actual runtime evidence from logs

Read the full file on GitHub · 142 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

5d ago First seen · 142 lines · 47 tokens per session scan A f318cb58cbd6

Subscribe to this mod's changes

bug-hunter is an agent published in the GitHub repository dykyi-roman/awesome-claude-code (96 stars, last pushed 19d ago), licensed MIT. It adds 47 tokens to every session and 1,039 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Code review Security

bug-hunter