security_auditor — Claude Code subagent: what it does, how to install
security_auditor — a subagent for Claude Code runs defensive application security reviews. Inspects source code, configurations, and dependency scanner logs to report exploitable flaws.
no license ★ 776 updated 2026-11-12
.claude/agents/security_auditor.mdWhat security_auditor does
This agent inspects supplied source fragments, infrastructure configs, scanner logs, and architecture boundaries. It returns a structured Markdown audit detailing each finding alongside prerequisites, plausible impact, remediation instructions, and unassessed attack surface.
Analysis builds on routing hints for tools like tool.lib.detect_secrets.v1, tool.bandit_security_scanner.v1, and Trail of Bits workflow skills. The agent operates under strict boundaries: fabricating CVE identifiers is disallowed, and scanner flags require contextual evidence before classification as exploitable.
What it helps with
- audit source code and configuration files for security flaws
- filter false positives from raw vulnerability scanner reports
- generate a structured remediation report with verification steps
Example prompts for the agent
- Audit the payment processing service and review its dependency scan
- Examine this repository configuration for exposed secrets and access flaws
- Review our API gateway route handlers and produce a risk assessment
How to install security_auditor
Manually: put the files in place
.claude/agents/security_auditor.mdFor all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention
Manually: put the files in place
.codex/agents/security_auditor.tomlCodex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/
Manually: put the files in place
.cursor/agents/security_auditor.mdCursor also reads .claude/agents/
Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean
Questions about security_auditor
How to install security_auditor in Claude Code?
Place the agent definition into .claude/agents/security_auditor.md.
What does security_auditor do?
It reviews supplied codebase assets, identifies architectural gaps or misconfigurations, and delivers an evidence-backed remediation report.
What rules govern security_auditor reviews?
The agent strictly conducts defensive validation, prohibits fabricated CVE records, and rejects marking findings as exploitable without concrete proof.