impact-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute shell commands such as
grepand directory listings to trace code references across the filesystem. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from project tickets, PRDs, and source code files, which could theoretically contain malicious instructions.
- Ingestion points: Accesses data from ticket files, PRD documents,
repo-map.md, and source code across multiple sibling directories. - Boundary markers: The instructions do not define specific delimiters or warnings to prevent the agent from obeying instructions potentially embedded in the analyzed files.
- Capability inventory: The agent can traverse the filesystem outside project roots and execute search-based shell commands.
- Sanitization: No explicit sanitization or validation of the data ingested from tickets or codebases is performed before inclusion in the final report.
Audit Metadata