Salesforce used January 2026 to take Slackbot from assistant to agent. The January 13 GA release gave Slackbot workspace search across conversations, files, threads, and DMs; document summarization; email drafting; and meeting scheduling. By March, it gained desktop awareness — the ability to observe and act across deals, conversations, calendar events, and declared habits outside the Slack window itself.
Salesforce also announced 30 new Slackbot features rolling out through mid-2026. That is a meaningful pace for a surface where many people run their entire working day.
What changed in Q1 2026
January: Agentic Slackbot goes generally available. Core capabilities:
- Workspace search across all Slack content (conversations, files, threads, canvases)
- Document upload for summarization and key extraction
- Email drafting and meeting scheduling from within Slack
February: Two additions that change the composability picture:
- Reusable AI-skills — custom task templates you define once and call via commands (e.g. "create a weekly review," "summarize this week's habit check-ins")
- MCP client support — Slackbot can now connect to external services via the Model Context Protocol, including Salesforce Agentforce integrations
March/April: Desktop awareness. Slackbot now operates outside Slack with visibility into calendar state, active tasks, and what Salesforce is calling "habit" context. The specific habits it tracks are tied to Agentforce's CRM use cases — sales cadences, follow-up rituals — but the underlying mechanism is the same one personal behavior agents use.
Why this matters if you use Slack for habits and routines
Slack is already where many people work. The friction to open a separate app for a habit check-in is real. If Slackbot can surface a morning habit nudge inside Slack, that's one fewer context switch.
The MCP client is the most significant shift for behavior agent users. MCP (Model Context Protocol) allows Slack to act as a client for external tool servers. A habit agent running its own MCP server can now push structured context into Slack natively — habit status, reminder triggers, briefing data — without relying on a custom Slack app integration.
The reusable AI-skills feature is interesting for teams running shared routines. A team standup routine or end-of-week review can be defined once as a Slackbot skill and invoked by anyone on the team with a short command.
What Slackbot still doesn't provide
The agentic Slackbot is a workspace assistant — it operates on Slack's content and Salesforce's CRM context. It does not:
- Maintain a dedicated habit model (cadence, streaks, exceptions, tolerance windows)
- Send reminders across channels outside Slack (Telegram, ChatGPT, browser)
- Adapt reminder timing based on missed behavioral patterns over time
- Give you a daily briefing that spans personal habits and team routines in one view
The distinction is between a workspace agent (Slackbot) and a personal behavior agent (Buffy). Both are useful; they operate on different data and serve different accountability functions.
How Buffy works with Slack
Buffy has a Slack channel adapter as one of several interfaces to the same behavior engine. Set up once — habits, tasks, routines, reminder timing — and Buffy reaches you in Slack, Telegram, ChatGPT, or OpenClaw depending on where you are.
For teams, this means Buffy reminders arrive in Slack alongside the rest of your work communication, while the behavior engine also follows individuals across other channels outside work hours.
The Slackbot expansion and Buffy are complementary, not redundant. Slackbot handles workspace productivity context. Buffy handles behavioral continuity across channels and time.
For teams already on Slack: Slack habit tracking and team setup walks through the practical configuration.
Next step
Get started with Buffy Agent →