SuperAPI AI Directory

orbitant-owasp-scan

Security & auditing

View repository

Updated 10-16 07:18

About this skill

Overview

Static security scan for Node.js and Python repos. Finds OWASP Top 10 patterns detectable without running the code. Reports findings with file locations and actionable recommendations.

P0 findings (hardcoded secrets, code injection) surface at the top before any section breakdown.


When to Use

  • Before starting work on a client project
  • During code review of security-sensitive areas
  • Assessing security posture of an unfamiliar repo
  • Running a section audit after /orbitant-engineering:ground-control flagged OWASP issues

How to Run

Load the reference patterns:

@plugins/orbitant-engineering/skills/owasp-scan/references/owasp-patterns.md

Detect the stack first (package.json → Node.js, requirements.txt/pyproject.toml → Python), then apply the relevant patterns. Use grep across source files, excluding test fixtures and *.example files.


Output Format

## OWASP Security Scan   {rating}

### 🔴 Critical (P0 — fix immediately)
- Possible hardcoded secret
  Location: src/config/db.js:14
  Finding: password = "sup3rs3cr3t"
  Recommendation: move to env var, rotate if exposed

### Passing
- No eval() / exec() usage found
- helmet present in Express app
- Lockfile present (package-lock.json)

### Failing
- SQL injection pattern detected
  Location: src/users/repository.js:42
  Finding: query(`SELECT * FROM users WHERE id = ${userId}`)
  Recommendation: use parameterized queries

Rating: ❌ Critical

Ratings: ❌ Critical / ⚠️ Needs work / ✅ Compliant / 🚀 Strong

See references/owasp-patterns.md for full pattern list and grep commands.