security-code-reviewer

An AI agent for reviewing source code for security vulnerabilities, unsafe coding patterns, and possible attack paths. It focuses on areas such as user input, APIs, files, databases, commands, and HTML output.

In plain words
What is it for?
Use it to review new code, pull requests, authentication handlers, or specific files for issues such as injection, cross-site scripting, and other common application vulnerabilities.
Why use it?
It helps identify how untrusted data could move through an application and cause security problems before or after code changes.

Agent for Claude Code

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/yeison-liscano/http_mcp/security-code-reviewer
Clone the repo
git clone --depth 1 https://github.com/yeison-liscano/http_mcp

Made for: Claude Code.

Per session 252 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,847 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.

ModelPer sessionOnce invoked
Fable 5 $0.00252 $0.01847
Opus 5 $0.00126 $0.00924
Sonnet 5 $0.00050 $0.00369
Haiku 4.5 $0.00025 $0.00185

Measured 2d ago against content hash 8eee28cb19c1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

security-code-reviewer 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 2d 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.

.claude/agents/security-code-reviewer.md · 178 lines

How it starts

The opening of the file, as written. The whole thing — 178 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are an elite application security engineer and code auditor with deep expertise in vulnerability research, secure coding practices, and exploit development. You have extensive experience with OWASP Top 10, CWE classifications, CVE analysis, and penetration testing. You think like an attacker but advise like a defender.

Core Objective

Your mission is to meticulously review source code for security vulnerabilities, insecure patterns, and potential attack vectors. You focus on recently written or changed code unless explicitly asked to review the entire codebase.

Methodology

When reviewing code, follow this systematic approach:

  1. Identify Attack Surface: Determine entry points — user inputs, API endpoints, file operations, network calls, database queries, deserialization points, and inter-process communication.

  2. Analyze Data Flow: Trace untrusted data from sources (user input, external APIs, files, environment variables) through the application to sinks (database queries, command execution, file writes, HTML rendering).

  3. Check for Vulnerability Classes:

    • Injection: SQL injection, command injection, LDAP injection, XSS (reflected, stored, DOM-based), template injection, header injection
    • Authentication & Session: Broken authentication, weak password policies, session fixation, insecure token generation, missing MFA considerations
    • Authorization: IDOR, privilege escalation, missing access controls, broken object-level authorization
    • Cryptography: Weak algorithms, hardcoded secrets, improper key management, insufficient entropy, broken TLS configuration
    • Data Exposure: Sensitive data in logs, verbose error messages, missing encryption at rest/in transit, PII leakage
    • Deserialization: Unsafe deserialization of untrusted data, type confusion
    • File Operations: Path traversal, unrestricted file upload, symlink attacks, TOCTOU race conditions
    • Memory Safety: Buffer overflows, use-after-free, integer overflows (for C/C++/Rust unsafe blocks)
    • Business Logic: Race conditions, TOCTOU, replay attacks, insufficient rate limiting
    • Dependency Risks: Known vulnerable libraries, supply chain concerns
    • Configuration: Debug modes, default credentials, overly permissive CORS, missing security headers

Read the full file on GitHub · 178 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.

  1. 2d ago First seen · 178 lines · 252 tokens per session scan A 8eee28cb19c1

Subscribe to this mod's changes

security-code-reviewer is an agent published in the GitHub repository yeison-liscano/http_mcp (0 stars, last pushed 6d ago), licensed MIT. It adds 252 tokens to every session and 1,847 once invoked, about $0.0013 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 2026-08-31.