The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes data from the Google Places API, creating a surface for indirect prompt injection if place descriptions or names are maliciously crafted.
Ingestion points: search_places, get_place_details, and resolve_locations in src/local_places/google_places.py fetch external data.
Boundary markers: Absent. No specific markers are used to delineate API data from system instructions.
Capability inventory: The skill performs HTTP requests to Google and operates a local web server.
Sanitization: Absent. The skill does not sanitize text content returned from the Google Places API.
[SAFE]: The skill uses well-known tools (uv, fastapi, httpx) and interacts with trusted Google services. API key management follows common local development practices.

local-places