tidas-bilingual-transcreation

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes the tiangong-lca CLI tool for data processing tasks including extraction and validation. This is a legitimate use of a vendor-owned utility for its defined purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data fields from JSONL files and other evidence sources.
  • Ingestion points: Data enters the context via processes.jsonl, trans-units.jsonl, and external source evidence files referenced in SKILL.md.
  • Boundary markers: No explicit delimiter-based safety instructions are present to separate untrusted data from processing logic.
  • Capability inventory: The agent can execute tiangong-lca commands and read/write to the local filesystem.
  • Sanitization: The skill utilizes a dataset bilingual validate command as a post-processing check to ensure the integrity and schema compliance of translated rows.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 08:31 PM
Security Audit — agent-trust-hub — tidas-bilingual-transcreation