The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the Google BigQuery public patent dataset, including patent titles, abstracts, and claims. Maliciously crafted content within these fields could lead to indirect prompt injection attacks. 1. Ingestion points: Data retrieved via search_patents and get_patent calls in SKILL.md. 2. Boundary markers: The instructions do not define delimiters or specific markers to isolate external patent content from agent instructions. 3. Capability inventory: The agent has access to Bash (for execution) and Write (for file system modifications) tools. 4. Sanitization: There is no documented logic for sanitizing or filtering the external patent content before it is processed by the agent.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute internal Python search scripts and setup commands like gcloud auth and python scripts/test_bigquery.py.
[EXTERNAL_DOWNLOADS]: The skill documentation mentions the installation of the google-cloud-bigquery library from the official package registry to facilitate interactions with Google Cloud services.

bigquery-patent-search