skills/mabhub/skills/calibre-library/Gen Agent Trust Hub

calibre-library

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes subprocess.run to execute the pgrep utility to check for active Calibre processes, ensuring the library database is not locked. It also uses standard Calibre command-line tools such as calibredb and calibre-debug, alongside system utilities like zip -d and gio trash, for ebook library maintenance and file management tasks.\n- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external and untrusted data by parsing EPUB metadata and internal structures, as well as referencing Wikipedia for title casing rules, which introduces a potential attack surface for indirect injection.\n
  • Ingestion points: The skill reads metadata from internal EPUB components (OPF files, containers) and local database entries.\n
  • Boundary markers: A strict operational cycle is enforced where the agent must provide a diagnostic report and recommendations, requiring an explicit user "ok" before any changes are applied.\n
  • Capability inventory: The skill is capable of performing write operations on the local library database and ebook files, including modifying archives and deleting files.\n
  • Sanitization: Implements normalization for metadata and provides specific instructions for the agent to identify and remove sensitive buyer identifiers (PII) without ever outputting their values to the agent's display.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 04:24 PM
Security Audit — agent-trust-hub — calibre-library