The Agent Skills Directory

Data Exposure & Exfiltration (LOW): The scripts discover_bases.py and query_records.py access the local .env file to retrieve the AIRTABLE_API_KEY. While this is the intended configuration mechanism, the access to sensitive files like .env is noted.
Indirect Prompt Injection (LOW): This skill ingests and processes untrusted data from the Airtable API. Ingestion points: Untrusted data enters via API responses in query_records.py (record fields) and discover_bases.py (metadata). Boundary markers: None are implemented to distinguish external content from instructions. Capability inventory: The skill uses requests.get for network communication and yaml.dump to write data to local cache files. Sanitization: No explicit sanitization or validation of external field content is performed before it is added to the agent's context.

airtable-master