Skills sast-bandit

📦

sast-bandit

Name: sast-bandit
Author: AgentSecOps

v0.1.0 Content revision r2 High Risk ⚡ Contains scripts⚙️ External commands📁 Filesystem access🌐 Network access🔑 Env variables

Scan Python Code with Bandit

Python security flaws can reach production unnoticed. This skill guides Bandit scans, prioritization, CI integration, and remediation with CWE and OWASP references.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "sast-bandit" from https://skillstore.io/skills/agentsecops-sast-bandit.md and its manifest at https://skillstore.io/api/skills/agentsecops-sast-bandit/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/agentsecops-sast-bandit.md Signed manifest GET /api/skills/agentsecops-sast-bandit/manifest Signed lockfile GET /api/skills/agentsecops-sast-bandit/lockfile

Test it

Using "sast-bandit". Review Bandit finding B602 in payments/tasks.py.

Expected outcome:

Priority: High because user input reaches a shell-enabled subprocess.
Recommended action: use a fixed executable with validated arguments and no shell.
Validation: test metacharacters, traversal values, and unexpected option-like inputs.

Using "sast-bandit". Plan a Bandit gate for a legacy Python service.

Expected outcome:

Start by recording all findings without blocking builds.
Block new high-severity, high-confidence findings after the baseline is reviewed.
Store restricted reports and require documented approval for suppressions.

Security Audit

High Risk

v10 • 7/23/2026 Open versioned report

Most static matches are false positives from configuration lists, Markdown formatting, references, and labeled vulnerable examples. Confirmed risks include unsafe eval suppression, unvalidated subprocess guidance, fragile xargs filename handling, and unsafe pickle suppression. Third-party hooks and actions also use mutable references.

Files scanned

1,527

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (2)

High

SKILL.md:240-246

Unsafe Pickle Suppression Guidance

The skill presents pickle.load on a caller-provided path as an acceptable # nosec example. A writable cache or attacker-controlled path could enable code execution.

The example directly deserializes a caller-selected file and suppresses B301 without enforcing a trusted directory, ownership, or integrity check.

Medium

assets/pre-commit-config.yaml:20-40 assets/pre-commit-config.yaml:100-145 SKILL.md:149-154

Mutable Third-Party Execution References

Pre-commit hooks and GitHub Actions execute third-party code through release tags instead of immutable commit hashes. A moved or compromised tag could execute altered code.

The cited configurations use version tags for code that runs in developer or CI environments. No immutable commit SHA or package hash is specified.

Capability review items (3)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

High

assets/bandit_config.yaml:204

Dynamic code evaluation with eval()

# result = eval(safe_expression) # nosec B307

The line is non-executable, but it presents eval() with a # nosec suppression and no validation. Copying this guidance can introduce arbitrary code execution.

High

references/remediation_guide.md:187

Python subprocess.run

subprocess.run(["cat", filename], check=True, capture_output=True)

The example labels this call safe although filename comes directly from request.GET. An attacker can select arbitrary readable paths or inject options into cat.

Medium

SKILL.md:274

xargs command (can execute arbitrary commands)

3. Scan only changed files in CI/CD: `git diff --name-only origin/main | grep '.py$' | xargs bandit`

Repository-controlled filenames are passed through newline and whitespace parsing into Bandit arguments. Crafted paths can split arguments or be interpreted as Bandit options.

Risk Factors

⚡ Contains scripts (2)

assets/bandit_config.yaml:204 references/cwe_owasp_mapping.md:61

⚙️ External commands (50)

assets/bandit_config.yaml:117 assets/bandit_config.yaml:112 assets/bandit_config.yaml:116 assets/bandit_config.yaml:112 assets/bandit_config.yaml:115 references/remediation_guide.md:187 references/remediation_guide.md:204 references/remediation_guide.md:209 references/remediation_guide.md:228 references/remediation_guide.md:169 references/remediation_guide.md:176 references/remediation_guide.md:172 SKILL.md:34-49 SKILL.md:49-57 SKILL.md:57-59 SKILL.md:59-61 SKILL.md:61 SKILL.md:63-79 SKILL.md:79-85 SKILL.md:85-97 SKILL.md:97-110 SKILL.md:110-115 SKILL.md:115-128 SKILL.md:128-132 SKILL.md:132 SKILL.md:140-159 SKILL.md:159-161 SKILL.md:161-165 SKILL.md:165-173 SKILL.md:173-177 SKILL.md:177-179 SKILL.md:179-181 SKILL.md:181-183 SKILL.md:183-185 SKILL.md:185-187 SKILL.md:187-189 SKILL.md:189-191 SKILL.md:191-199 SKILL.md:199-206 SKILL.md:206-212 SKILL.md:212-219 SKILL.md:219-225 SKILL.md:225-234 SKILL.md:234-240 SKILL.md:240-247 SKILL.md:247-251 SKILL.md:251-257 SKILL.md:257-264 SKILL.md:264-265 SKILL.md:265-266

📁 Filesystem access (15)

assets/bandit_config.yaml:8 assets/bandit_config.yaml:10 assets/bandit_config.yaml:12 assets/bandit_config.yaml:19 assets/bandit_config.yaml:20 assets/bandit_config.yaml:23 assets/bandit_config.yaml:27 assets/bandit_config.yaml:28 assets/bandit_config.yaml:29 assets/bandit_config.yaml:30 assets/bandit_config.yaml:31 assets/bandit_config.yaml:75 references/cwe_owasp_mapping.md:23 SKILL.md:68 SKILL.md:272

🌐 Network access (29)

🔑 Env variables (19)

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/agentsecops-sast-bandit/audits/10?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/agentsecops-sast-bandit/security.svg)](https://skillstore.io/skills/agentsecops-sast-bandit?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/agentsecops-sast-bandit?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/agentsecops-sast-bandit/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/agentsecops-sast-bandit.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

AgentSecOps. (2026). sast-bandit security audit report (audit version 10) [Author version 0.1.0]. Skillstore. https://skillstore.io/skills/agentsecops-sast-bandit/audits/10

BibTeX citation

@techreport{agentsecops-agentsecops-sast-bandit-2026,
  author = {AgentSecOps},
  title = {sast-bandit security audit report (audit version 10)},
  institution = {Skillstore},
  year = {2026},
  number = {10},
  url = {https://skillstore.io/skills/agentsecops-sast-bandit/audits/10},
  note = {Author version 0.1.0}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "sast-bandit security audit report (audit version 10)"
version: "0.1.0"
type: report
authors:
  - name: "AgentSecOps"
date-released: "2026-07-23"
url: "https://skillstore.io/skills/agentsecops-sast-bandit/audits/10"
identifiers:
  - type: other
    value: "skillstore:agentsecops-sast-bandit:audit:10"
    description: "Skillstore immutable audit report identifier"

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

100

Maintainability

Content

Community

Spec Compliance

What You Can Build

Triage a Python Feature

Scan changed Python files, review high-confidence findings, and prepare focused fixes before code review.

Establish an Application Baseline

Inventory existing Bandit findings, map them to CWE and OWASP, and define remediation priorities.

Add a CI Security Gate

Configure Bandit or pre-commit checks that block selected severity levels while preserving reviewable reports.

Try These Prompts

Run a Basic Scan

Scan the Python files in [path] with Bandit. Use a read-only workflow, explain the command, and summarize findings by severity and confidence.

Review a Bandit Report

Review the Bandit report at [path]. Prioritize exploitable findings, identify likely false positives, and recommend one verification step for each priority item.

Design a CI Gate

Design a Bandit gate for [repository]. Specify scan scope, thresholds, exclusions, report retention, failure behavior, and a safe exception review process.

Build a Remediation Plan

Analyze Bandit findings in [report]. Map each confirmed issue to CWE and OWASP, rank exploitability, propose fixes, and define validation tests.

Best Practices

Run scans with read-only source access and restrict report access.
Review exploitability before suppression, and document every accepted exception.
Pin Bandit, hooks, and CI actions to reviewed immutable versions.

Avoid

Do not treat every Bandit finding as a confirmed vulnerability.
Do not suppress eval, pickle, or shell findings without a documented trust boundary.
Do not publish reports containing credentials or sensitive source snippets.