cursor-mcp-auth

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides configuration details (URL and UUID) for a specific Atlassian Jira instance. These are public identifiers required for the Cursor MCP integration to function correctly and do not constitute sensitive data exposure.
  • [SAFE]: The instructions include security best practices, explicitly forbidding the agent from requesting or printing OAuth secrets and directing the user to use the built-in browser-based authentication flow.
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates tools that ingest data from external Jira issues, which is a potential surface for indirect prompt injection.
  • Ingestion points: Jira issue summaries and descriptions via the getJiraIssue tool.
  • Capability inventory: Reading and searching Jira data.
  • Boundary markers: Not specified in this authentication-focused skill.
  • Sanitization: Not specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 08:39 AM
Security Audit — agent-trust-hub — cursor-mcp-auth