Kilo Code, an open-source AI coding startup backed by GitLab’s former CEO Sid Sijbrandij, has released a Slack integration that allows software engineers to execute code changes and submit pull requests without ever leaving their team’s chat. This move addresses a core friction point in modern development: the constant switching between tools and loss of context.
The launch comes as the AI-assisted coding market is rapidly expanding, with major acquisitions and funding rounds demonstrating its growing importance. But unlike many competitors focused on siloed coding assistants, Kilo is betting on embedding AI directly into existing workflows where decisions are actually made.
“Engineering teams don’t make decisions in IDE sidebars; they make them in Slack,” explains Kilo Code co-founder and CEO Scott Breitenother. “This bot lets you do everything—and more—without ever leaving the chat.”
Why This Matters: The Rise of “Vibe Coding”
The integration targets a trend known as “vibe coding,” where large language models are used to write and modify code based on conversational context. This approach is gaining traction because it streamlines development by eliminating the need to re-explain problems in separate tools. Microsoft now reports that AI-generated code accounts for 30% of its codebase, while Google recently acquired an AI coding startup for $2.4 billion.
Kilo’s approach directly addresses this trend by making AI-assisted coding accessible within the primary communication hub for many engineering teams: Slack.
How Kilo for Slack Works
The integration operates by allowing users to mention @Kilo in Slack threads. The bot then reads the entire conversation, accesses connected GitHub repositories, and either answers questions about the codebase or automatically creates a branch and submits a pull request.
For example, if a product manager reports a bug in Slack, engineers can discuss the issue and then simply type: “@Kilo based on this thread, can you implement the fix for the null pointer exception in the Authentication service?” The bot handles the rest, spinning up a cloud agent, implementing the fix, and submitting a pull request directly within Slack.
Kilo’s Competitive Edge
Kilo explicitly positions itself against leading AI coding tools like Cursor and Claude Code. The company argues that Cursor’s Slack integration is limited to single repositories, while Claude Code lacks persistent conversational context for longer workflows.
Kilo claims its integration works across multiple repositories simultaneously, maintains context across extended Slack threads, and enables seamless handoffs between Slack, IDEs, cloud agents, and the command line.
Model Choice and Security Concerns
Perhaps the most notable aspect of the launch is Kilo’s decision to default to MiniMax’s M2.1 model, a Chinese AI company that recently completed an IPO in Hong Kong. This move has raised questions about enterprise security, given concerns over sending proprietary code through Chinese infrastructure.
Breitenother addresses this directly: “MiniMax’s IPO drew backing from major global investors, indicating strong confidence in their models.” He emphasizes that MiniMax models are hosted by U.S.-compliant cloud providers like AWS, Google, and Microsoft.
Kilo also stresses that it is model-agnostic, allowing enterprise customers to choose their preferred models based on security and compliance requirements.
The Future of AI-Assisted Development
Kilo’s launch reflects a shift in the AI coding market toward integration rather than standalone tools. The company believes the winners will be those who can seamlessly embed AI into existing workflows, meeting developers where they already work.
Whether a 34-person startup can outmaneuver industry giants like OpenAI and Anthropic remains to be seen. However, if Kilo is right that the hard problem isn’t generating code but integrating it, the company may have found a sustainable competitive advantage.
