Reporting false-positive safety blocks on benign custom MCP actions

🌐 链接: https://community.openai.com/t/reporting-false-positive-safety-blocks-on-benign-custom-mcp-actions/1406376
🏷️ 分组: OpenAI Developer Community
🕒 时间: 2026-10-15 02:00:05
OpenAI Developer Community
Reporting false-positive safety blocks on benign custom MCP actions
Reporting false-positive safety blocks on benign custom MCP actions
Environment Product surface: ChatGPT web with a user-authorized custom MCP integration. Scope: non-production, offline validation and ordinary document work. Account details, machine identifiers, local paths, credentials, request IDs and internal project…