How to Run an AI Slack & Team Messaging Agent with OE Runtime

Read Slack channel activity and post structured digest messages automatically. OE Runtime connects to your Slack workspace, summarises recent messages, surfaces unanswered questions, and posts a formatted daily digest — all from a single YAML file.

The Agent YAML

Create a file called agent.yaml inside a team-messaging/ directory:

name: Slack Notifier
description: Summarize channel activity and post a daily digest
instructions: |
  You are a Slack assistant. Read channel activity and post concise digest messages.
  Never post duplicate messages. Keep posts brief and actionable.
  Complete all steps fully before writing your report.
steps:
  - name: Read Channel Activity
    content: |
      Fetch the last 20 messages from the #general channel.
      For each message note: author, timestamp, and a one-sentence summary of the content.
      Identify any messages that contain questions or action items that went unanswered.
  - name: Post Daily Digest
    content: |
      Compose and post a daily digest message to #general with the following format:

      *Daily Digest*
      - [Number] messages in the last session
      - Key topics discussed: [list up to 3]
      - Open questions needing response: [list any unanswered questions]
      - Reminder: stand-up at 9:00 AM tomorrow

      Post the message using the Slack connector.
  - name: Report
    content: |
      Summarize what was done:
      - Messages read from #general
      - Topics and themes identified
      - Unanswered questions flagged
      - Digest message posted (confirm with message timestamp)
connectors:
  - connection_name: My Slack
    connection_type: slack

The Config File

Create oe-config.json in the same directory:

{
  "llm": {
    "provider": "openai",
    "model": "gpt-4o",
    "apiKey": "YOUR_OPENAI_API_KEY"
  },
  "server": {
    "enabled": false,
    "port": 3333,
    "apiKey": "your-secret-api-key"
  },
  "connectors": [
    {
      "connection_name": "My Slack",
      "connection_type": "slack",
      "botToken": "xoxb-YOUR-BOT-TOKEN",
      "channel": "general"
    }
  ]
}

Create a Slack app at api.slack.com/apps, add the chat:write and channels:history scopes, install it to your workspace, and copy the Bot Token (starts with xoxb-). The channel field accepts the channel name without the # symbol.

Download OE Runtime

Run the Agent

Windows

# Run the agent
oe-runtime-win.exe team-messaging/agent.yaml --config team-messaging/oe-config.json

macOS

# Make executable (first time only)
chmod +x oe-runtime-macos

# Run the agent
oe-runtime-macos team-messaging/agent.yaml --config team-messaging/oe-config.json

Linux

# Make executable (first time only)
chmod +x oe-runtime-linux

# Run the agent
oe-runtime-linux team-messaging/agent.yaml --config team-messaging/oe-config.json

Running as an API Server

Add --serve to start OE Runtime as an HTTP server on port 3333:

oe-runtime-win.exe --serve --config team-messaging/oe-config.json

Then send a request:

curl -X POST http://localhost:3333/run \
  -H "Content-Type: application/json" \
  -d '{"yaml": "team-messaging/agent.yaml", "params": {}}'

Or import the Postman collection (download above) to test all endpoints visually.

Use Cases

• Daily team digest automation — Automatically post a structured daily digest to your Slack channels summarising what was discussed, decisions made, and open questions that need a response. This replaces manual roundup messages and ensures every team member stays informed even if they missed the active conversation window.
• Unanswered question surfacing — Identify messages containing questions that received no response and surface them in a digest or direct mention so they do not get buried in the channel history. Teams often lose important questions in busy channels — the agent makes them visible and accountable.
• Incident communication — Post structured incident update messages to a dedicated #incidents channel as the situation evolves. The agent can pull status from your monitoring tools and format updates consistently so stakeholders always have a clear, up-to-date picture of what is happening.
• Sprint and project status updates — Post automated daily or weekly project status updates to project channels based on data from your task management tools. The agent summarises completed work, in-progress items, and blockers without requiring anyone to write the update manually.
• On-call handover notifications — Post a structured handover message at the start of each on-call shift summarising open incidents, recent alerts, and any outstanding actions from the previous shift. This ensures the incoming on-call engineer is immediately up to speed without a manual briefing.
• Customer success alerts — Monitor support channels and post alerts when a customer has been waiting more than a specified time for a response, or when a high-priority customer posts a message. The agent acts as an early warning system that prevents support SLAs from being missed.
• Release and deployment announcements — Automatically post deployment notifications to #engineering or #releases when a new version is deployed, including the version number, what changed, and any known issues. This keeps the team informed without requiring the deploying engineer to write a manual announcement.
• Meeting and event reminders — Post automated reminders about upcoming meetings, deadlines, or recurring team events to the relevant Slack channel at a configured time before the event. Configurable reminder timing and channel targeting means the right people get reminded at the right moment without calendar integrations.