wb-preview

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to construct and run shell commands by interpolating user-provided variables, including a 'Correction' filter string. The instruction to 'accept a raw filter string verbatim' without specifying sanitization or escaping mechanisms creates a surface where a malicious user could provide input containing shell metacharacters (e.g., semicolons or backticks) to execute unauthorized commands.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: User-supplied values for 'Correction' filter spec, 'Source video', and 'Timestamp' found in the SKILL.md procedures.
  • Boundary markers: Absent. The skill does not instruct the agent to use delimiters or 'ignore' warnings for the interpolated user content.
  • Capability inventory: The skill uses Bash(ffmpeg), Bash(ffprobe), and file writing capabilities.
  • Sanitization: Absent. There are no instructions provided to the agent to validate, escape, or sanitize the user-provided filter string before it is executed in the shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 03:34 PM
Security Audit — agent-trust-hub — wb-preview