oma-translator

Pass

Audited by Gen Agent Trust Hub on Mar 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, direct prompt injections, or unauthorized data access were found in the skill files.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external strings for translation. Ingestion points: Source text analysis and locale file scanning (SKILL.md). Boundary markers: None explicitly defined to separate instructions from content. Capability inventory: Accessing project-specific files and providing text output (SKILL.md). Sanitization: Mitigation through rules for preserving placeholders and flagging ambiguity (SKILL.md). This surface is consistent with the skill's primary purpose.
  • [METADATA_POISONING]: Metadata fields are descriptive and contain no hidden instructions or deceptive content.
  • [SAFE]: Execution protocol references appear to be part of the standard agent platform environment and do not involve untrusted remote code.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 21, 2026, 09:36 PM