accelint-persona-review
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill utilizes standard platform integration patterns, specifically the Figma and Outline MCP servers, to perform its primary function. Analysis of all skill files and referenced persona profiles shows no evidence of hardcoded credentials, malicious command execution, or persistence mechanisms. All network-related activity is mediated through official MCP tools.
- [INDIRECT_PROMPT_INJECTION]: The skill maintains a data ingestion surface for external content.
- Ingestion points: Designs are fetched via the Figma MCP tool
mcp__figma-desktop__get_design_context, and supporting documentation is retrieved via the Outline MCPListMcpResourcesTool(SKILL.md). - Boundary markers: The instructions do not provide explicit delimiters or "ignore embedded instructions" warnings for the external data retrieved from Figma or Outline.
- Capability inventory: The agent has the capability to read local persona files, retrieve Figma design contexts, and search/read Outline documents.
- Sanitization: There is no documented sanitization or validation of the data retrieved from external platforms before it is analyzed by the agent.
Audit Metadata