project-real-estate-agent
Pass
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill instructions create a surface for indirect prompt injection by requiring the agent to perform live web searches for current mortgage rates and property trends. This ingestion of untrusted external content could allow embedded instructions in third-party websites to influence agent responses.
- Ingestion points: External websites accessed via the 'WebSearch' command as described in the Trend Awareness Protocol.
- Boundary markers: The prompt does not specify the use of delimiters or 'ignore' instructions when processing search results.
- Capability inventory: The skill leverages the agent's WebSearch capability; no scripts, file-system access, or network exfiltration commands are present in the provided files.
- Sanitization: There are no mechanisms described for sanitizing, escaping, or validating the data retrieved from external web sources.
Audit Metadata