Workato Slack Genies Make AI Tool Runs Less Invisible
AI agents running inside Slack have a visibility problem. When a Genie — Workato’s term for AI agents built in Agent Studio — processes a request, users have had little way to see what the agent is actually doing between the moment they ask a question and the moment it responds. That gap creates a specific kind of friction: users either distrust the output because they cannot trace it, or they interrupt the agent mid-run because they assume nothing is happening. Workato’s June 2026 update to Slack Genies addresses this directly with intermediate messages and a persistent tool call stream.
What intermediate messages and persistent tool call feedback mean in practice
The update introduces four specific changes to how Genies surface their work inside Slack.
Intermediate messages are always shown. Any response-relevant content the underlying language model returns during processing is now surfaced in Slack and in Test Mode. Previously, intermediate outputs might not appear at all — users saw the final response but not the reasoning or status updates that led there. Now those messages are visible in sequence, giving users a real-time view of what the agent intends to communicate at each step.
Persistent tool call stream. Alongside intermediate messages, a real-time stream shows the specific actions the Genie takes in response to a query. This stream runs continuously while the agent is processing and persists after it completes. That persistence is the meaningful part: the thread now contains an observable record of what the agent did, not just what it said.
Collapsible streams. Users can collapse any tool call stream to reduce visual clutter in a Slack thread. The stream keeps running in the background and can be expanded again at any time. For users who do not care about step-by-step visibility, this keeps the thread readable. For users who need to audit what happened, the full detail is still there when needed.
Human-readable tool labels. Internal tool calls now display plain-language labels instead of raw function names. An action that previously showed as a technical identifier now shows as something like “Searching knowledge base [My KB] for [query]…”. Teams do not need to understand the underlying tool architecture to read what the agent did.
The workflow benefit for small teams
For small teams using Workato Genies in Slack, the persistent tool call stream turns Slack threads into a lightweight operational record. If a Genie runs a knowledge base search, pulls a file, and then synthesizes a response, that sequence is now documented in the thread itself. The next team member who looks at the thread can see not just the answer but how the agent arrived at it.
This matters most for shared workflows — support queues, internal operations channels, or anywhere multiple people interact with the same agent across different sessions. When something goes wrong or a response seems off, the tool call stream gives the team a starting point for diagnosing what happened without escalating to an administrator or pulling logs.
The collapsible design respects the fact that most of the time, most users just want the answer. The visibility is opt-in per interaction — expand when you need it, collapse when you do not.
What to check before rolling out broadly
Before expanding Genie usage in Slack on the basis of this update, a few things are worth checking across your setup.
Privacy and data exposure in shared channels. Tool call streams now show query text in plain language inside Slack — for example, “Searching knowledge base [My KB] for [specific query text].” If users are querying sensitive information, that query text now appears in the Slack thread. Confirm that the channels where Genies are deployed have appropriate membership restrictions before turning this on broadly.
Channel clutter in high-volume channels. Even with collapsible streams, intermediate messages add turn-by-turn content to threads. In channels where Genies are invoked frequently, this increases thread length significantly. Test the visual impact in a pilot channel before rolling out to high-traffic spaces.
Test Mode parity. The update explicitly states that intermediate messages appear in both Slack and Test Mode. If your team validates Genie behavior in Test Mode before production deployment, confirm that the test environment now reflects the same step-by-step output that users will see in live Slack threads.
Permissions configuration. Human-readable labels reduce the opacity of what agents are accessing, but they do not change the underlying access controls. Review what data sources and tools each Genie has access to separately from this update. Visibility into what an agent does is not a substitute for scoping what it is allowed to do.
Who this matters for, who can ignore it
This update is relevant for Workato customers using Genies as live tools in Slack, particularly teams where more than one person interacts with an agent in the same channel or thread. The combination of intermediate messages, persistent tool call history, and readable labels directly reduces the oversight gap that makes teams hesitant to rely on AI agents for real operational work.
If your team uses Workato but has not deployed Genies in Slack, this update does not affect your current workflows. If you are evaluating whether to enable Slack Genies for your team, the observability improvements make the case for a pilot stronger — you are not handing the channel over to a black box, and users can see exactly what actions each agent takes.
Teams not on Workato or not using agentic automation in Slack have nothing to act on here.
Source: Workato official changelog, “Intermediate Messages and Persistent Tool Call Feedback — Slack,” June 2026. Additional product documentation at docs.workato.com/en/agentic/agent-studio.