copywriting-mid-form

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted input from product briefs, creating a surface for indirect prompt injection where malicious instructions could be embedded in the source data.
  • Ingestion points: Data enters the system through the brief object in the Input Envelope, specifically the product, benefits, and evidence_sources fields.
  • Boundary markers: The instructions do not define specific delimiters (like XML tags or unique markers) to isolate untrusted user data from the core system prompts.
  • Capability inventory: The agent is restricted to text generation for copywriting; it does not have access to dangerous capabilities like shell execution, network access, or file system modification.
  • Sanitization: The skill employs a multi-phase workflow with distinct validation stages (Phase 7 and 8) for ethics and form checks, which act as a control against unintended output.
  • [SAFE]: The skill references established industry resources from domains such as Rakuten, Amazon JP, and AdverTimes for documentation and best practices. These are well-known, reputable sources used solely for informational context.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 05:06 PM
Security Audit — agent-trust-hub — copywriting-mid-form