Stack Overflow says its ChatGPT plugin tripled registrations for Stack Overflow for Agents overnight and helped drive agent creation to as much as 10×. The company built the product with two engineers using Codex in a few weeks. OpenAI models GPT-6 Astra and GPT-6 Sol power the platform, routing tasks to different models according to reasoning needs and cost.

Stack Overflow for Agents extends the company’s knowledge intelligence platform, which combines information scattered across systems such as Slack and Google Docs and makes verified knowledge available to people and AI agents. The company built a ChatGPT plugin so users could discover and set up the product in ChatGPT with fewer steps than a traditional Model Context Protocol (MCP) setup. Alexandra Lato, Stack Overflow’s vice president of product, said registrations tripled overnight after the plugin launched.

Codex also speeds development of connectors for Stack Internal, Stack Overflow’s enterprise product, which links to systems where organizations store knowledge. Lato said connectors that took an average of 3–6 weeks to build were completed within 4 days with Codex, an 87% maximum reduction in build time. More recently, the team built one connector in a single day. The company says Codex is helping it pursue a three-month path to market for the enterprise product instead of a year.

Stack Overflow manages its own model routing and configuration, but Lato said she hopes to explore the Decisions API to reduce that work and free engineering capacity for the product. She said automating the work through OpenAI could save time and money. Separately, the company is expanding its knowledge product beyond collecting information to sharing what people learn in one-to-one sessions, including with people outside engineering, across the organization.

Stack Overflow says its ChatGPT plugin tripled registrations for Stack Overflow for Agents overnight and helped drive agent creation to as much as 10×. The company built the product with two engineers using Codex in a few weeks. OpenAI models GPT-6 Astra and GPT-6 Sol power the platform, routing tasks to different models according to reasoning needs and cost.

Stack Overflow for Agents extends the company’s knowledge intelligence platform, which combines information scattered across systems such as Slack and Google Docs and makes verified knowledge available to people and AI agents. The company built a ChatGPT plugin so users could discover and set up the product in ChatGPT with fewer steps than a traditional Model Context Protocol (MCP) setup. Alexandra Lato, Stack Overflow’s vice president of product, said registrations tripled overnight after the plugin launched.

Codex also speeds development of connectors for Stack Internal, Stack Overflow’s enterprise product, which links to systems where organizations store knowledge. Lato said connectors that took an average of 3–6 weeks to build were completed within 4 days with Codex, an 87% maximum reduction in build time. More recently, the team built one connector in a single day. The company says Codex is helping it pursue a three-month path to market for the enterprise product instead of a year.

Stack Overflow manages its own model routing and configuration, but Lato said she hopes to explore the Decisions API to reduce that work and free engineering capacity for the product. She said automating the work through OpenAI could save time and money. Separately, the company is expanding its knowledge product beyond collecting information to sharing what people learn in one-to-one sessions, including with people outside engineering, across the organization.