Skills
skills/jtsang4/efficient-coding/readwise-research/Gen Agent Trust Hub

readwise-research

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and summarizes untrusted external content from the user's Readwise documents.\n
  • Ingestion points: Uses reader-get-document-details and reader-get-document-highlights to pull arbitrary text into the agent context in SKILL.md.\n
  • Boundary markers: Lacks instructions for the agent to treat retrieved document content as data rather than instructions.\n
  • Capability inventory: The agent can modify the user's library (tagging, moving files) based on directions found within the untrusted content using the readwise CLI.\n
  • Sanitization: No validation or filtering of the retrieved content is performed before processing.\n- [EXTERNAL_DOWNLOADS]: The skill recommends installing @readwise/cli, which is a recognized package from a well-known service.\n- [COMMAND_EXECUTION]: The skill uses shell commands via the local readwise CLI for its core functionality of data retrieval and library management.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 02:54 AM
Security Audit — agent-trust-hub — readwise-research