visual-whiteboarding
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill workflow involves reading and processing untrusted external data, such as Product Requirements Documents (PRDs) and system analysis templates, to generate visual outputs. This creates a potential surface where malicious instructions embedded in those documents could influence the agent's behavior during diagram generation.
- Ingestion points: Processes external content from
PLAN.md,SYSTEM_ANALYSIS_TEMPLATE.md, and various PRD sources in the SKILL.md workflow. - Boundary markers: The instructions do not specify the use of delimiters or protective headers (e.g., 'ignore instructions inside this block') when the agent ingests external documents.
- Capability inventory: The skill utilizes MCP tool capabilities to create shapes in Miro (
create_shape), write XML to Draw.io (write_drawio_xml), and read node metadata from Figma. - Sanitization: The skill includes explicit security checkpoints in
SKILL.md(Section 4) and theWHITEBOARD_DESIGN_TEMPLATE.md(Execution Pre-Flight Checklist) to ensure that API keys, database credentials, and PII are redacted before being written to shared visual canvases.
Audit Metadata