The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes functional requirement files via the $ARGUMENTS parameter to generate design documentation. This creates a vulnerability surface where a malicious input file could contain instructions designed to trick the agent into performing unintended actions.
Ingestion points: Step 1 reads the functional requirements file provided as an argument.
Boundary markers: There are no explicit boundary markers or instructions to disregard embedded commands in the input file.
Capability inventory: The skill uses Write and Edit tools which could be misused if the agent obeys instructions within the untrusted requirements file.
Sanitization: No sanitization or content validation steps are mentioned for the requirements document.
[DYNAMIC_CONTEXT_INJECTION]: The skill uses shell execution to generate dynamic paths for document storage.
Evidence: The syntax !date +%Y-%m-%d is used in Step 3, 4, and 5 to create timestamped folders.
Assessment: This is a benign and legitimate use of the date utility to ensure documentation is correctly timestamped and organized, posing no significant security risk as it does not incorporate external input into the shell command.

design