connect-notes
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data (user notes) which could theoretically contain malicious instructions.
- Ingestion points: User-supplied notes provided via paste, folder selection, or search results as described in the 'What to ask for' section.
- Boundary markers: Absent; the instructions do not specify delimiters to separate note content from the agent's instructions.
- Capability inventory: The agent has the ability to read, search, and create/modify notes through external connectors.
- Sanitization: The instructions do not mandate sanitizing or escaping the text found within notes.
- Mitigation: The skill design enforces a strict 'Propose, then wait' protocol, ensuring no changes to the user's data occur without explicit manual confirmation, which significantly mitigates risks associated with untrusted input.
- [NO_CODE]: The skill is composed entirely of Markdown instructions and does not include any scripts, dependencies, or compiled binaries.
Audit Metadata