seven-prompt-content-engine
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill ingests untrusted user-provided data, including text ideas, writing samples, and content from external URLs. These inputs are interpolated into prompts via placeholders (e.g., [WRITING_SAMPLE]) in the referenced templates. The instructions do not include boundary markers, delimiters, or explicit sanitization to prevent embedded instructions within user data from influencing agent behavior. * Ingestion points: SKILL.md (Intake table) and references/prompt-templates.md (Placeholders). * Boundary markers: Absent. * Capability inventory: Reading user-provided URLs. * Sanitization: Absent.
- [EXTERNAL_DOWNLOADS]: The skill instructions in SKILL.md direct the agent to read content from user-provided URLs to inform the drafting process. This involves fetching external data into the agent's context.
Audit Metadata