surprise-me

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it analyzes user-generated highlights and document metadata without specific boundary markers. • Ingestion points: Readwise highlights and document records (SKILL.md). • Boundary markers: None; the agent is not instructed to ignore commands embedded in user data. • Capability inventory: Read-only access via MCP and vendor CLI. • Sanitization: No sanitization is performed on highlight text.
  • [COMMAND_EXECUTION]: Directs the agent to use the readwise CLI for data retrieval. This is a trusted vendor resource from the author readwiseio used for its intended purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 04:22 PM