adobe-design-from-template
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements a dynamic instruction mechanism via the
importantNotefield in tool responses. The agent is instructed to prioritize these notes for formatting and next-step actions, which represents an indirect prompt injection surface. - Ingestion points: Tool outputs from
search_design,fill_text,replace_image,change_background_color, andanimate_design(SKILL.md). - Boundary markers: None explicitly defined to separate external data from instructions within the
importantNotefield. - Capability inventory: The skill uses tools to modify designs and generate pre-signed PDF download URLs (
download_design). - Sanitization: No specific sanitization or validation of the
importantNotecontent is described before the agent processes it. - [SAFE]: All tools (
adobe_mandatory_init,search_design, etc.) and external resource patterns align with the vendor's (Adobe) infrastructure and the skill's intended purpose for visual design editing.
Audit Metadata