webapp-testing

Fail

Audited by Snyk on Mar 10, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.60). The prompt instructs a dispatched subagent to read a local/internal file ("~/.claude/skills/webapp-testing/SKILL.md"), an action outside normal web-app testing needs that could access/internalize sensitive data, so it constitutes a hidden/deceptive instruction.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.90). The SKILL.md workflow explicitly tells the agent to navigate to and interact with arbitrary web pages (e.g., page.goto('https://example.com'), "Screenshot a URL", and the "Reconnaissance-Then-Action" steps that "identify selectors from rendered state" and "Read page content"), so untrusted third‑party page content can be fetched, interpreted, and used to drive automation decisions.
Audit Metadata
Risk Level
CRITICAL
Analyzed
Mar 10, 2026, 03:57 AM