The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes curl commands to interact with a Calibre library's web API. These operations are intended to search for book records and retrieve metadata details for the filtering process.
[EXTERNAL_DOWNLOADS]: The skill retrieves book metadata from a remote API endpoint. This access is necessary for the skill to perform its stated purpose of auditing external library contents.
[DATA_EXFILTRATION]: No unauthorized data transmission was detected. The skill processes metadata retrieved from the user's library and generates a summary report within the agent's context.
[PROMPT_INJECTION]: The skill processes untrusted metadata such as book comments and titles, which presents a surface for indirect prompt injection. However, the skill implements explicit filtering logic via regular expressions and string matching, which mitigates the risk of the agent misinterpreting processed data as instructions.

calibre-book-screening