evernote
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from Evernote notes, which are untrusted external sources.
- Ingestion points: External data is ingested via the
evernote call-toolcommand when performing search or read operations (e.g.,notes.read). - Boundary markers: The instructions do not define specific delimiters or isolation prompts to separate note content from the agent's control flow.
- Capability inventory: The skill uses a binary
evernotetool with capabilities to read, search, and create notes based on user permissions. - Sanitization: There are no documented steps for sanitizing or filtering the content retrieved from notes.
Audit Metadata