skills/zircote/human-voice/voice/Gen Agent Trust Hub

voice

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted writing samples and open-ended responses through internal agents and an NLP pipeline, creating a surface for indirect prompt injection attacks.
  • Ingestion points: Writing samples stored in ~/.human-voice/writing_samples/ and open-ended conversational responses are ingested into the agent context.
  • Boundary markers: The instructions do not define boundary markers or delimiters to differentiate between system instructions and user-provided data.
  • Capability inventory: The skill has access to Bash, Write, and specialized agent orchestration.
  • Sanitization: The skill does not describe any sanitization or validation logic for the ingested samples.
  • [COMMAND_EXECUTION]: The skill uses the Bash tool and explicitly ignores platform-prescribed data storage directories to maintain its own persistence and profile data under ~/.human-voice/.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 06:43 PM
Security Audit — agent-trust-hub — voice