oz-report-pr

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a shell command using an environment-defined variable $OZ_CLI to process pull request metadata.
  • [INDIRECT_PROMPT_INJECTION]: The skill's command template includes placeholders for pull request URLs and branch names that are directly interpolated into the shell command. This creates a surface for indirect prompt injection (command injection) if the source data is untrusted.
  • Ingestion points: The <url> and <branch> placeholders in SKILL.md are intended to be filled with repository-specific metadata.
  • Boundary markers: The skill uses single quotes (') to wrap the variables in the shell command, which provides a basic but insufficient layer of protection against maliciously crafted strings.
  • Capability inventory: The skill utilizes shell execution capabilities to call the $OZ_CLI binary.
  • Sanitization: No input validation or filtering is performed on the data before it is passed to the shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 04:59 AM
Security Audit — agent-trust-hub — oz-report-pr