cfe-diff
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of 1C source code files (.bsl) and XML configuration files, which could theoretically contain malicious instructions for the agent.
- Ingestion points: The script reads file contents from the user-specified
ExtensionPathandConfigPathdirectories (specificallyConfiguration.xmland various.bslmodules). - Boundary markers: The output uses structural markers (e.g.,
[BORROWED],[TRANSFERRED],[OWN]) to help the agent distinguish between analysis results and raw source code. - Capability inventory: The skill is limited to reading files (
Read,Glob) and executing a local analysis script. It does not have network access or the ability to modify files outside its scope. - Sanitization: No specific sanitization or escaping is performed on the code blocks extracted from the modules before they are output to the agent's context.
Audit Metadata