understand-diff
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
git diffandgrep. These are standard tools for version control and text searching within the local repository and do not pose a high security risk in this context. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from the codebase knowledge graph.
- Ingestion points: The agent reads from
.understand-anything/knowledge-graph.jsonto perform its analysis. - Boundary markers: There are no explicit instructions to the agent to ignore or delimit instructions found within the knowledge graph content.
- Capability inventory: The skill has the ability to execute shell commands (
git,grep) and write to local files (.understand-anything/diff-overlay.json). - Sanitization: There is no mention of sanitizing or validating the contents of the knowledge graph before processing.
Audit Metadata