jianying-export-prep

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified. The skill instructions prioritize transparency and safety by requiring auditing of plans, execution, and evidence.
  • [COMMAND_EXECUTION]: The skill relies on the execution of a local jianying CLI. It mitigates risks by explicitly forbidding automatic updates or installations, ensuring the execution environment remains under user control.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a defined attack surface via the ingestion of project drafts and media files.
  • Ingestion points: Processes project media, drafts, and configuration files as specified in SKILL.md (Step 3).
  • Boundary markers: Implements structural validation (project verify) and clearly defines the differences between proxy and native output to prevent confusion.
  • Capability inventory: Performs CLI-based file verification, rendering, and metadata probing.
  • Sanitization: Uses SHA-256 hash verification for inputs and outputs, enforces strict schema validation for job files, and mandates isolated output paths to prevent original data corruption.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 11:55 PM
Security Audit — agent-trust-hub — jianying-export-prep