The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data—including test specifications, handoff documents, and user requirements—to generate automation scripts, creating a surface for indirect prompt injection.
Ingestion points: The skill consumes test-spec files, Testany Automation Handoff data, and user natural language input as defined in SKILL.md.
Capability inventory: The agent is capable of generating executable scripts for Python, Playwright, and Java, as well as creating ZIP archives.
Boundary markers: No explicit delimiters or instructions are provided to the agent to isolate or ignore potentially malicious instructions embedded within the imported specifications.
Sanitization: The instructions do not define sanitization or escaping protocols for external content before it is interpolated into the code generation templates.
[DYNAMIC_EXECUTION]: The skill generates executable script files at runtime based on user-provided logic and requirements.
Evidence: Predefined templates in references/executors/ for PyRes, Playwright, Maven, and Postman are used to dynamically assemble script files like test_api.py and example.spec.ts.
Logic: The agent generates these scripts from templates by incorporating logic derived from external test requirements, which is a low-severity dynamic code generation pattern.

testany-case-writing