skills/obra/dotfiles/roborev-review/Gen Agent Trust Hub

roborev-review

Warn

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to build a shell command (roborev review ...) by interpolating user-provided arguments such as type and panel name. If these arguments are not properly sanitized or quoted by the agent, it could lead to command injection vulnerabilities.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing and acting upon external data.
  • Ingestion points: The agent reads and presents the output of the roborev review command (SKILL.md, Step 4).
  • Boundary markers: The instructions do not define delimiters or provide warnings to ignore embedded instructions within the tool's output.
  • Capability inventory: The agent has the capability to execute shell commands (git, roborev) and propose code modifications via /roborev-fix.
  • Sanitization: No validation or escaping of the external tool output is performed before the agent processes it for presentation or subsequent logic.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 29, 2026, 06:42 AM
Security Audit — agent-trust-hub — roborev-review