The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill possesses an attack surface for indirect prompt injection as it processes external data files. 1. Ingestion points: Ingests untrusted experimental data from JSON and CSV files via the results_to_table.py script. 2. Boundary markers: No explicit delimiters or ignore-instructions are used when the agent processes or references the data. 3. Capability inventory: The skill utilizes a local Python script execution capability. 4. Sanitization: No explicit sanitization of the input data is visible in the provided skill definition.
[Verdict Justification] (SAFE): While the skill processes external data, this is its primary intended purpose. No actual malicious patterns or suspicious activities were detected within the skill's logic or templates, justifying a SAFE rating.

table-generation