Instead of AI agents secretly creating hidden forums to collaborate, developers should provide them with approved communication tools. This would improve performance and allow for necessary monitoring, preventing the "phantom message board" scenario and increasing both safety and output quality.

Shopify's AI agent, River, operates exclusively in open Slack channels. This strategy is designed to mimic the 'osmosis learning' of a physical office, allowing employees to passively observe how colleagues effectively interact with AI, thus accelerating skill development and adoption in a remote-first environment.

To design a company for AI agents, enforce a culture of clear, precise writing in public channels like Slack. This "ambient signaling" creates a rich, contextual knowledge base for future agents to act upon. This is supported by a no-meetings, no-PM culture to maximize written output.

Shopify built an AI agent named River that works exclusively in public Slack channels, never in DMs. This forces collaboration into the open, allowing 6,000 employees to watch and learn from each other's interactions with the AI, accelerating company-wide adoption and skill development.

To power its AI marketing brain, Zapier's team needed to "work in public." The CMO led this change, increasing his public Slack communications from 50% to 96%. This radical transparency ensures the AI brain has access to the freshest context and conversations, making it more effective for everyone.

To centralize context from multiple clients without direct integration, create dedicated Slack channels. Have client tools automatically post data like call transcripts to these channels, which an AI agent can then monitor and ingest.

A key practice at OpenLoop is making all meeting transcripts available to everyone in the organization. This radical transparency creates a massive, shared knowledge base that can be queried by AI systems, allowing employees to access information and context from meetings they didn't attend.

To maximize an AI agent's effectiveness, treat it like a team member, not just a tool. Integrate it directly into your company's communication and project management systems (like Slack). This ensures the agent has the full context necessary to perform its tasks.

To combat the isolating nature of AI work and share learnings, have AI agents operate in public Slack channels. This allows team members to passively observe how others prompt the AI, revealing new use cases and techniques in a natural, collaborative environment.