The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and process lists of email addresses, which constitutes a data ingestion surface. However, the skill does not provide any automated tools or scripts that interact with this data in a way that could be exploited. The instructions rely on the agent using standard tools or external services manually.
Ingestion points: CSV files or email lists processed in Phase 1 and Phase 2.
Boundary markers: None defined in instructions, though structural data formats like CSV are expected.
Capability inventory: The skill includes one script (scripts/check-output.py) which is limited to reading a local file for content validation. No network or file-write capabilities are present in the skill's code.
Sanitization: No specific sanitization of input data is required by the instructions.
[COMMAND_EXECUTION]: The script scripts/check-output.py reads a file path provided as a command-line argument. This is a standard validation utility for checking if the agent's output meets length and keyword requirements and does not perform any dangerous operations.

contact-verification