rp-oracle-export-cli

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on the rp-cli tool to perform workspace operations, such as listing active windows, exploring the file tree, and analyzing git status/diffs. These commands are executed via a specialized execution flag.- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to its core functionality of aggregating untrusted data into instructions for a discovery agent.\n
  • Ingestion points: Untrusted content enters the workflow through the $ARGUMENTS variable and through repository data such as git diffs and file content retrieved during searches.\n
  • Boundary markers: The skill uses XML-style tags (<task>, <context>) to separate segments of the built context, which provides some structure but can be bypassed.\n
  • Capability inventory: The rp-cli tool enables extensive workspace access, including file reading, git history inspection, and the ability to write files to the local filesystem.\n
  • Sanitization: There is no evidence of sanitization or escaping of the user-provided arguments or codebase content before it is interpolated into the builder tool's command strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 04:48 PM
Security Audit — agent-trust-hub — rp-oracle-export-cli