The Agent Skills Directory

[PROMPT_INJECTION] (HIGH): The skill is vulnerable to indirect prompt injection due to its ability to read external data and perform actions based on it. \n
Ingestion points: The list_events function in scripts/google_calendar.py ingests event data (summaries, descriptions) from the Google Calendar API. \n
Boundary markers: Absent. The skill does not provide delimiters or instructions to the agent to treat calendar content as untrusted data. \n
Capability inventory: The create_event function in scripts/google_calendar.py allows the agent to perform write operations (creating events). \n
Sanitization: Absent. No filtering or sanitization is performed on meeting data before it is presented to the agent. \n- [DATA_EXFILTRATION] (HIGH): The script accesses sensitive credential files in the user's home directory. \n
Evidence: The script accesses credentials.json and token.json within the ~/.calendar_credentials/ directory. These files contain sensitive OAuth 2.0 secrets and tokens that provide access to the user's Google account data. While used for legitimate API communication, this represents high-severity sensitive path access.

google-calendar