record-a-decision
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes an
exectool to perform maintenance tasks including listing directory contents (ls), searching for keywords within files (grep), and reading file contents (cat). These shell operations are limited to the project's documentation directories (decisions/andproposals/). - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from local markdown files, which presents a surface for indirect prompt injection if those files were to contain malicious instructions.
- Ingestion points: Data is read from the filesystem using
exec("cat ...")andsearch()operations targeting thedecisions/andproposals/folders. - Boundary markers: The instructions do not specify the use of delimiters or dedicated markers to separate ingested file content from the agent's instruction context.
- Capability inventory: The skill has access to tools for creating files (
write), modifying existing documents (edit), and executing local shell utilities (exec). - Sanitization: There is no explicit requirement for the agent to sanitize or validate the content retrieved from documentation files before processing it.
Audit Metadata