sales-push-notification
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements a self-improvement loop through the
references/learnings.mdfile. It is instructed to read this file at the start of every session and append new information discovered during the conversation. This creates a surface for Indirect Prompt Injection (Category 8). - Ingestion points: The agent gathers context from the user regarding push notification strategies and technical setups.
- Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands when writing to or reading from the
learnings.mdfile. - Capability inventory: The skill has the capability to write to the local file system (
references/learnings.md). - Sanitization: No sanitization or validation of the content being appended to the learnings file is specified.
Audit Metadata