email-sequence

Pass

Audited by Gen Agent Trust Hub on Apr 5, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructs the agent to read marketing and product context from a local file (".claude/product-marketing-context.md") to tailor the email sequences. This behavior creates a surface for indirect prompt injection, as any instructions maliciously embedded in that context file could influence the agent's behavior.
  • Ingestion points: Loading of ".claude/product-marketing-context.md" in the Initial Assessment section of SKILL.md.
  • Boundary markers: Absent; there are no specified delimiters or instructions to ignore commands within the ingested content.
  • Capability inventory: High-level text generation and strategic planning for automated email communications.
  • Sanitization: Absent; the skill does not define validation or filtering steps for the data retrieved from the context file.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 5, 2026, 08:06 PM
Security Audit — agent-trust-hub — email-sequence