The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a collection of Python scripts in the scripts/ directory (e.g., batch_convert_templates.py, convert_formats.py) that utilize subprocess.call to execute other internal scripts for parsing and converting QA output formats. These scripts are used to process project templates and artifacts.
[DATA_EXPOSURE]: Conversion scripts perform read and write operations on the local file system, specifically targeting the output-templates/ and artifacts/ directories. While intended for processing QA data, this capability allows the agent to interact with files on the host system.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data, such as requirement documents, user stories, and application logs provided by the user. This creates a surface where malicious instructions embedded in that data could influence the agent's behavior during analysis.
Ingestion points: External documents and logs provided by the user at runtime for analysis (referenced in prompts/requirements-analysis_EN.md and others).
Boundary markers: Prompts use horizontal dividers to separate instructions from user data, but do not implement strict sanitization or "ignore" directives for embedded content.
Capability inventory: The skill can read/write local files and execute Python scripts via the scripts/ directory.
Sanitization: No explicit sanitization or filtering of external data is performed before it is included in the prompt context.

daily-testing-workflow-en