social-proof-architect

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill is composed entirely of natural language instructions and psychological frameworks. It does not include any scripts, shell commands, or tool invocations that could pose a technical risk.
  • [NO_CODE]: There are no code files, package dependencies, or automation scripts associated with this skill. It functions as a specialized persona and instruction set for the AI agent.
  • [PROMPT_INJECTION]: The instructions emphasize ethical boundaries and honest representation of data. There are no patterns suggesting attempts to bypass safety filters or exfiltrate system prompts.
  • [INDIRECT_PROMPT_INJECTION]: While the skill processes external input (user-provided marketing context), it lacks the capabilities (network access, file writing, or command execution) required to exploit such an injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 07:16 AM
Security Audit — agent-trust-hub — social-proof-architect