StablyAI/agent-slack provides a command-line interface (CLI) for automating Slack interactions, specifically designed for AI agents. It addresses the need for efficient data retrieval and manipulation within Slack, optimizing output for cost-effective consumption by Large Language Models (LLMs) through token-efficient JSON formatting. This tool benefits AI developers and users by simplifying Slack automation and reducing LLM processing costs.

How It Works

Built with TypeScript and Bun, agent-slack prioritizes token efficiency by outputting compact JSON with pruned fields and minimal duplication. Its core advantage lies in its zero-configuration authentication, which seamlessly integrates with Slack Desktop on macOS, with fallbacks for browser-based authentication or manual token imports. This approach eliminates complex setup, allowing agents to interact with Slack immediately.

Quick Start & Requirements

Installation is straightforward via Bun: curl -fsSL https://raw.githubusercontent.com/stablyai/agent-slack/master/install.sh | sh. Alternatively, use npm: npm i -g agent-slack (requires Node >= 22.5). Zero-config authentication works out-of-the-box on macOS by reading Slack Desktop data. Fallbacks include Chrome extraction or manual imports via agent-slack auth import-*. Environment variables (SLACK_TOKEN, SLACK_COOKIE_D) or standard Slack tokens (xoxb/xoxp) can also be used.

Highlighted Details

• Token-Efficient Output: Aggressively pruned JSON, removing null/empty fields for LLM cost savings.
• Automated File Handling: Snippets, images, and attachments are auto-downloaded to local paths for agent consumption.
• Comprehensive Slack Operations: Supports fetching messages and threads, searching messages and files (with filters), sending replies, adding reactions, and converting Slack Canvases to Markdown.
• Agent Skill Integration: Ships with an agent skill compatible with platforms like Claude Code, Codex, and Cursor, installable via skills.sh or manual scripts.

Maintenance & Community

No specific details regarding maintainers, community channels (like Discord/Slack), or roadmap were provided in the README.

Licensing & Compatibility

The README does not specify a software license. Compatibility for commercial use or closed-source linking is undetermined without a license.

Limitations & Caveats

The zero-configuration authentication mechanism is primarily optimized for macOS environments. Resolving channel names across multiple Slack workspaces requires explicit use of the --workspace flag.