security

name	security
description	Run a quick security audit on recent changes or specific files

Security Skill

Invokes the security-auditor agent for quick security checks on code.

Information Needed

1. Target - What to audit:
   • Recent changes (default)
   • Specific file or directory
   • Specific module (crypto, network, auth)

Steps

1. Determine audit scope:

   • If no target specified, audit recent git changes
   • If file/directory specified, focus on that
   • If module specified, audit that module

2. Launch security-auditor agent via Task tool:

Task({
  subagent_type: "security-auditor",
  prompt: "Perform a security audit on [target]. Focus on:
    - Input validation
    - Authentication/authorization
    - Cryptographic usage
    - IPC security (Tauri commands)
    - OWASP Top 10 vulnerabilities

    Provide findings with severity ratings and remediation guidance."
})

3. Review agent findings and present summary

Focus Areas

Area	What to Check
Input Validation	User input sanitization, type checking
Auth Flows	Skeleton Key handling, session management
Crypto	XChaCha20-Poly1305 usage, key derivation
IPC Security	Tauri command permissions, invoke patterns
P2P Sync	mDNS discovery, TCP connections, encryption

Example: Audit Recent Changes

User: "Run a security check"

1. Get recent changes:

git diff HEAD~5 --name-only

2. Launch security-auditor agent with file list

3. Present findings summary

Example: Audit Crypto Module

User: "Security audit the crypto module"

Launch agent:

Task({
  subagent_type: "security-auditor",
  prompt: "Perform a security audit on src-tauri/src/crypto/ focusing on:
    - BIP39 mnemonic handling
    - HKDF key derivation
    - XChaCha20-Poly1305 encryption
    - Key storage security
    - Potential side-channel vulnerabilities"
})

Notes

• Always runs via the security-auditor agent (model: opus)
• Findings include severity ratings (Critical, High, Medium, Low)
• CWE classifications provided where applicable
• Remediation guidance included for each finding