The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the calibredb command-line utility to perform library operations. It includes a helper script scripts/list_books_json.py which uses the Python subprocess module to call the tool. The script follows security best practices by passing arguments as a list rather than a shell string, which prevents command injection.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes metadata from external ebook files. A malicious actor could craft an ebook with metadata (such as the title or author) containing instructions intended to override the agent's behavior.
Ingestion points: scripts/list_books_json.py, calibredb list, and calibredb show_metadata commands.
Boundary markers: Absent. The agent is instructed to process the output of these tools directly without specific instructions to ignore embedded commands.
Capability inventory: The skill has the ability to read from and write to the filesystem, and to connect to remote content servers.
Sanitization: No sanitization or filtering of book metadata is performed before it is presented to the agent.
[DATA_EXFILTRATION]: The skill accesses local files (ebook libraries) and can connect to remote Calibre Content servers using user-provided URLs. While these are intended functions, they represent a data exposure surface. The skill includes a specific security warning advising users against passing passwords as literal command-line arguments to prevent them from appearing in process lists.

calibredb