Command & code injection

Why this agent matters

Command and code injection is usually game over: the attacker runs code as your application. The 2017 Equifax breach began with an injection flaw in Apache Struts.

Calls to subprocess or eval are common and often harmless — a fixed command, an argument list with no shell. This agent checks whether input is user-controlled and whether arguments are passed safely, so real sinks stand out.

What it checks

• os.system, subprocess with shell=True, popen and backticks
• eval / exec and dynamic code evaluation
• Server-side template injection
• Arguments built from request data

When it marks a finding as a likely false positive

• The command is fixed, or arguments are passed as a list without a shell
• Input is validated against an allow-list before use

The fix it suggests

• Argument vectors without shell=True
• Removing eval / exec
• Allow-list validation of any input that must reach a command

Scope and limits

This agent validates findings reported by Semgrep, Gitleaks and Trivy; it doesn't hunt for new bugs on its own. A finding reaches it when its CWE or rule/path keywords match this vulnerability class. It sees only the minimal code around the finding, with secrets redacted, and returns a verdict with a confidence score, a plain-English explanation and a suggested patch.

Common questions

Related agents

• SQL injection — Decides whether attacker-controlled input actually reaches a SQL or NoSQL query unparameterized.
• Cross-site scripting (XSS) — Checks whether untrusted data is rendered into HTML or JavaScript without contextual encoding.
• Server-side request forgery (SSRF) — Checks whether a server-side request can be pointed at a host the attacker chooses.

See all 20 review agents → · Agent setup reviewer · How the pipeline works