xrefs
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Analysis reveals a surface for indirect prompt injection due to the ingestion of untrusted data from binaries during analysis.
- Ingestion points: The skill queries string content (
string_refs), imported API names (imports), and function symbols from analyzed binaries (SKILL.md). - Boundary markers: No specific delimiters or boundary markers are utilized to isolate untrusted binary content from the agent's internal instruction set.
- Capability inventory: The skill is configured with access to Bash, Read, Glob, and Grep tools.
- Sanitization: There is no evidence of sanitization or validation performed on the strings or symbols retrieved from the binary file before they are processed by the agent.
Audit Metadata