The Agent Skills Directory

SAFE (SAFE): No security issues were detected across all analyzed files.
Prompt Injection: The content consists of instructional material for software testers. References to "SQL injection" and "XSS" in references/experience-based-techniques.md are correctly contextualized as examples of security defects to look for during testing, not as instructions to the agent.
Data Exposure & Exfiltration: No hardcoded credentials, API keys, or sensitive file paths were found. Templates (e.g., assets/templates/test-environment-checklist.md) use safe placeholders for configuration data.
Obfuscation: All documentation and templates are in clear text. No Base64, zero-width characters, or homoglyphs were detected.
Unverifiable Dependencies & RCE: While the skill references playwright and Node.js, it does not include any scripts that perform installations or execute remote code. Commands like npx playwright test are provided as documentation for manual or CI usage.
Indirect Prompt Injection (SAFE): The skill provides templates that interpolate data (e.g., {{feature}} in assets/templates/test-conditions.md). While this represents a data ingestion surface, the skill lacks any active logic or executable scripts to process this data, resulting in no risk of exploitation.
Ingestion points: assets/templates/*.md files containing {{variable}} placeholders.
Boundary markers: Absent in templates.
Capability inventory: No active scripts or command execution logic included.
Sanitization: No sanitization logic provided, but unnecessary given the lack of executable capabilities.

qa-manual-istqb