write-cold-message
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data without explicit security boundaries.
- Ingestion points: Recipient and offer details provided by the user (SKILL.md, Step 1) and outreach examples retrieved via
get_cold_message_examples(SKILL.md, Step 2). - Boundary markers: The instructions do not define delimiters (e.g., XML tags or triple quotes) to separate untrusted content, nor do they instruct the model to ignore potential commands embedded in the data.
- Capability inventory: The skill invokes the
get_cold_message_examplestool and references a secondary skillfind-buyersfor automated delivery. - Sanitization: No input validation or sanitization logic is present to filter malicious instructions from the processed text.
Audit Metadata