skills/wkentaro/agent-skills/hear-me/Gen Agent Trust Hub

hear-me

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [NO_CODE]: The skill consists entirely of configuration and instructional markdown; it does not include any executable scripts, binaries, or automated installation steps.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process user-supplied dictated text, which creates a surface for indirect prompt injection where malicious instructions could be embedded in the data.
  • Ingestion points: User messages (dictated requests) are ingested as arguments to the skill or as follow-up messages while the mode is enabled in SKILL.md.
  • Boundary markers: The skill does not specify any delimiters or boundary markers to isolate the user's dictated input from the instructions.
  • Capability inventory: The skill restricts model invocation within its own scope (disable-model-invocation: true) but influences the parameters of subsequent task work.
  • Sanitization: The protocol explicitly instructs the agent to replace sensitive values like API tokens with placeholders, which acts as a sanitization step for the output provided to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 03:29 AM
Security Audit — agent-trust-hub — hear-me