alterlab-cdm-subtitle-loc
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to read and build context from external, user-provided project files.\n
- Ingestion points: Workflow Step 1 (Spotting & Segmentation) explicitly mentions reading existing project files such as screenplays, dialogue lists, and translation notes.\n
- Boundary markers: There are no instructions defining boundaries or directing the agent to ignore potential instructions embedded within the source dialogue or project metadata.\n
- Capability inventory: The agent has file-writing capabilities (creating SRT and VTT files) and network access (web searching for standards).\n
- Sanitization: The instructions do not include steps for sanitizing, filtering, or validating the textual content extracted from project files.
Audit Metadata