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
- Identify the target — Determine the Python file(s) or directory to scan.
- 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)
- For specific tests:
- Parse the results — Read the JSON output and present findings:
| # | Severity | Confidence | Test ID | File:Line | Issue | Remediation |
|---|----------|------------|---------|-----------|-------|-------------|
- 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
Repository
vchirrav/produc…i-skillsGitHub Stars
1
First Seen
Feb 14, 2026
Security Audits
Installed on
opencode2
gemini-cli2
antigravity2
mistral-vibe2
claude-code2
github-copilot2