ChatGPT adds MCP Events support

// 58m agoPRODUCT UPDATE

ChatGPT adds MCP Events support

ChatGPT now supports the MCP Events spec, letting MCP servers push messages, comments, and status changes into connected sessions instead of waiting for a tool call. Developers can use those events to trigger user automations from live backend activity.

// ANALYSIS

This is the missing half of MCP: moving from request-response tools toward event-driven agents that can react as work happens.

–Enables workflows triggered by new messages, task completions, incidents, and status changes
–Reduces the need for inefficient polling and stale context
–Developers still need robust authentication, retries, ordering, deduplication, and consent controls
–Broader MCP event semantics remain experimental, so portability across clients is not guaranteed yet

// TAGS

chatgptmcpautomationagenttool-usestreaming

DISCOVERED

58m ago

2026-09-30

PUBLISHED

1h ago

2026-09-30

RELEVANCE

9/ 10

AUTHOR

mxstbr

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

RESEARCH26m ago

Graphite Maps 13,000 Shifting AI Tells

Graphite analyzed 10,000 human articles and 90,000 AI-generated articles across nine models, identifying nearly 13,000 words, phrases, and stylistic patterns associated with AI writing. The study finds that model-specific tells change between releases, while AI writing often resembles other AI writing more than human prose. [Research](https://graphite.io/five-percent/research/ai-tells)

OPEN SOURCE1h ago

Monid Open-Sources Agent Tool Router

Monid has open-sourced its MIT-licensed connector layer for routing AI agents across 2,000+ tools and 72+ providers. Its discover, inspect, and run workflow lets agents compare tool pricing, health, latency, and schemas before paying per call. [GitHub](https://github.com/monid-ai/monid) [Docs](https://docs.monid.ai/)

POLICY1h ago

AGMAI Issues Responsible AI Math Guidelines

The independent Advisory Group on Mathematics and Artificial Intelligence published recommendations for labs releasing mathematical results their researchers do not yet understand. It calls for public, versioned deposits, proper citations, model and process disclosures, formalization where possible, and funding for community-led human understanding.