security-diff-scan
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- Command Execution: The skill executes local Python scripts (
generate_in_scope_files.py,normalize_candidates.py,finalize_scan_contract.py) located within the plugin directory to manage scan scope and results. These are standard helper scripts for the skill's operation. - Indirect Prompt Injection: The skill is designed to ingest and analyze untrusted data (source code changes, pull requests, and URLs). While it treats these as untrusted, the analysis process itself constitutes an attack surface where malicious instructions embedded in code could attempt to influence the agent's review process.
- File System Access: Accesses local repository files and writes scan artifacts (e.g.,
threat_model.md,report.md,findings.json) to a context directory. This is consistent with its stated purpose of performing security audits.
Audit Metadata