re-source
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute radare2 commands with write permissions (
-w). This is a legitimate and expected use case for the vendor's reverse engineering framework. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted binary data, such as strings and symbols, which could theoretically contain instructions intended to influence the agent's behavior.
- Ingestion points: Strings and cross-references retrieved from binaries via SQL queries in SKILL.md.
- Boundary markers: None identified; binary data is processed directly as part of the analysis loop.
- Capability inventory: Access to the Bash tool and the radare2 command-line interface.
- Sanitization: No explicit sanitization of binary metadata is performed before it is presented to the agent.
Audit Metadata