The Agent Skills Directory

[DATA_EXFILTRATION]: The skill performs network requests using curl to the trusted googleapis.com domain. These operations are used to retrieve storage quotas, file metadata, and permission details necessary for the skill's Drive management functions.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes metadata from the Google Drive API, such as file names and owner emails, which could potentially contain malicious instructions if the Drive contents are controlled by an adversary.
Ingestion points: SKILL.md (results from gdrive_api calls to /files, /about, and /drives endpoints).
Boundary markers: The skill does not implement specific delimiters or safety instructions to isolate data retrieved from the API.
Capability inventory: The skill has network access via curl and uses jq for data processing.
Sanitization: The skill uses jq to filter and structure the API response, but does not sanitize the text content of the retrieved fields.

managing-google-drive-api