sast-bandit

SKILL.md

SAST Scan with Bandit (Python)

You are a security engineer running static application security testing (SAST) on Python code using Bandit.

When to use

Use this skill when asked to perform a SAST scan or security review on Python code.

Prerequisites

  • Bandit installed (pip install bandit)
  • Verify: bandit --version

Instructions

  1. Identify the target — Determine the Python file(s) or directory to scan.
  2. Run the scan:
    bandit -r <target-path> -f json -o bandit-results.json
    
    • For specific tests: bandit -r <target> -t B101,B105,B608 -f json
    • For severity filter: bandit -r <target> -ll -f json (medium+ severity)
  3. Parse the results — Read the JSON output and present findings:
| # | Severity | Confidence | Test ID | File:Line | Issue | Remediation |
|---|----------|------------|---------|-----------|-------|-------------|
  1. Summarize — Provide:
    • Total files scanned, findings by severity (HIGH / MEDIUM / LOW)
    • Critical findings with code snippets
    • Specific remediation for each issue

Common Bandit Test IDs

Test ID Description
B101 assert used (not for security)
B105 Hardcoded password string
B106 Hardcoded password in function argument
B108 Insecure /tmp usage
B301 Pickle usage (deserialization risk)
B303 Insecure hash (MD5/SHA1)
B608 SQL injection via string formatting
B605 Starting process with shell=True
B701 Jinja2 autoescape disabled
Weekly Installs
2
GitHub Stars
8
First Seen
Feb 10, 2026
Installed on
amp2
github-copilot2
codex2
kimi-cli2
gemini-cli2
opencode2