aiqbench analysis
Slack's Code API makes its "code channel" workspace available to third-party and in-house AI agents, so any developer can have an agent open a channel, add people and files, post progress, and archive it when done. For AI readers, the substance is the coordination layer for agentic coding: human oversight and approvals happen inside a shared chat space rather than a private agent session. The claims of scale rest on GitHub Copilot's reported figure that 30% of its agent sessions open a code channel, which is a vendor-reported number and not independently verified.
Introducing the Slack Code API: Your agents, your team, one code channel

Any custom agent can now join the team in Slack Code. The Slack Code API is generally available, so teams and their agents can build and ship code together.
In August, we launched Slack Code with a simple belief: agentic development shouldn’t be single-player. Instead of disappearing into a private tab, agent work should happen where your team already works, out in the open. Launch partners including Anthropic, Cognition, GitHub, and Vercel put that concept to the test. Code channels are what make this multiplayer AI concept work, providing a dedicated space that an agent can open for a task, with the right people, context, and plan already in it. Teammates can watch the agent work, add context, and approve changes without leaving Slack. GitHub Copilot is proving this works at scale: It opens a dedicated code channel on 30% of its agent sessions. That means nearly a third of the time, the work is big enough for a task-sized workspace where the team can go from idea to done. “We’re building…
Excerpt shown under the publisher's RSS terms · 143 words · full text and figures are on the publisher's site
Related coverage
Mixture of Channel Experts: Static Sparse Supports with Input-Adaptive Mixing for Pointwise Projections
arXiv · Aug 26, 2026
EMGBlend: Heterogeneity-Aware Self-Supervised Pretraining for Gesture and Force Decoding
arXiv · Sep 24, 2026
Forward-Only Convolutional Neural Networks with Learnable Channel-Class Assignment
arXiv · Jun 10, 2026
Channel-Oriented Design for EEG-to-Music Reconstruction
arXiv · Jun 4, 2026