The Agent Skills Directory

[COMMAND_EXECUTION]: The skill implements file system write capabilities through the export_to_excel method, which uses the pandas library and the openpyxl engine to generate validation reports. This behavior is consistent with the skill's stated purpose of providing facility handover data.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data as it processes BIM model elements, properties, and classifications provided by users. While this data is validated against regular expression patterns, the processing of external data constitutes a standard indirect injection attack surface.
Ingestion points: BIM elements list passed to the check_model method in SKILL.md.
Boundary markers: None explicitly implemented for data interpolation, though processing is limited to structured dictionary keys.
Capability inventory: File system write access via pandas.ExcelWriter.
Sanitization: Uses re.match for property validation, which restricts values to defined patterns but does not sanitize content before writing to Excel files.
[SAFE]: The skill implementation uses standard Python libraries for data science and validation. No evidence of prompt injection, obfuscation, credential harvesting, or unauthorized network activity was found. The logic aligns perfectly with the vendor's documented BIM validation business case.

ids-checker