Skills
Navigate
skills/faberlens/hardened-skills/walkie-talkie-hardened/Gen Agent Trust Hub

walkie-talkie-hardened

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local scripts and binaries (tools/transcribe_voice.sh and bin/sherpa-onnx-tts) to process audio and generate speech.
  • [EXTERNAL_DOWNLOADS]: The SAFETY.md documentation includes links to external vendor resources at faberlens.ai and GitHub for support and further evaluation.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface. Transcribed text from audio messages is processed as a standard user prompt without explicit boundary markers or sanitization, which could allow instructions hidden in audio to influence the agent.
  • Ingestion points: SKILL.md (Workflow Step 1).
  • Boundary markers: Absent; transcribed text is treated as a normal user prompt.
  • Capability inventory: Shell execution (tools/transcribe_voice.sh, bin/sherpa-onnx-tts) and messaging via the message tool.
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 06:31 PM
Security Audit — agent-trust-hub — walkie-talkie-hardened