review-delta
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the local codebase (git diffs and source code snippets) which could contain embedded instructions intended to influence the agent's behavior during a review. This is an inherent risk for code review tools.
- Ingestion points: Data is ingested via
get_review_context_tool()andquery_graph_tool()as described inSKILL.md. - Boundary markers: None identified in the prompt templates.
- Capability inventory: The skill uses tools for graph building (
build_or_update_graph_tool), querying (query_graph_tool), and documentation access (get_docs_section_tool). It does not appear to have arbitrary file write or network capabilities. - Sanitization: No specific sanitization or filtering of the ingested source code is described.
Audit Metadata