spectacles-ai

Pass

Audited by Gen Agent Trust Hub on Mar 28, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides examples of ingesting user speech via the AsrModule and passing the resulting transcript to LLMs. This establishes a surface for indirect prompt injection.
  • Ingestion points: session.onTranscriptUpdate event handlers in SKILL.md and references/rsg-asr-patterns.md.
  • Boundary markers: No boundary markers or 'ignore' instructions are included in the prompt templates.
  • Capability inventory: The skill possesses network capabilities (RemoteServiceModule), camera access (CameraModule), and audio output (TtsModule).
  • Sanitization: No input sanitization or validation is implemented in the provided snippets.
  • [DATA_EXFILTRATION]: The skill includes code for capturing camera frames and audio transcripts and transmitting them to remote AI endpoints. This behavior is the intended and documented primary use case for the skill's vision and speech processing capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 28, 2026, 04:22 PM