Back/Engineering/Notion/Notion
IntermediateEngineering

Automate Daily Standup Preparation with a Custom Notion AI Agent

Have a Notion Custom Agent assemble the daily standup pre-read from recent Slack discussion, Notion tasks, GitHub pull requests, meeting notes, and Honeycomb metrics, then post the page to your team channel.

Automate Daily Standup Preparation with a Custom Notion AI Agent

Ryan opens Notion's Hot Potato agent, shows its 9:00 AM schedule, fans work out across Honeycomb, Slack, tasks, and prior notes, then posts the pre-read to Slack.

Before you start

What you need

  • Notion Custom Agents and a recurring meeting-page template
  • A team Slack channel and Notion task database
  • GitHub access plus Honeycomb through MCP or equivalent telemetry access
  • A defined standup format and meeting time

What you’ll make

A fresh Notion standup page populated before the meeting with current metrics, progress, decisions, bugs, questions, and a Slack notification linking to it.

Tools used

Step by step

The workflow

Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.

5 steps

Step01

Automate Page Creation

Create a recurring Notion automation that makes one blank standup page from your meeting template before the agent runs. Include the meeting date, project, and team in properties so the agent can identify the correct page without guessing.

Step02

Configure the Custom Agent

Create a Notion Custom Agent and schedule it after the page automation. Ryan runs Hot Potato at 9:00 AM each day. Give it edit access to the meetings database, read access to the project sources, and permission to post the final link in the team Slack channel.

Step03

Connect Data Sources

Connect the exact Slack channel, task database, GitHub repository, prior meeting notes, and Honeycomb query. Ryan configured Honeycomb through MCP and gave the agent a screenshot of the working query. Test each source once before combining them.

Step04

Synthesize the Pre-Read

Tell the agent to gather the prior 24 hours of activity, fan the source searches out when subagents are available, and write into today's meeting page. Define the sections and ask for links back to source discussions, tasks, pull requests, and metrics.

Example prompt
Prepare today's standup pre-read in the meeting page created for [team or project]. Cover the previous 24 hours in [time zone]. Gather CI metrics from the saved Honeycomb MCP query, updates and questions from [Slack channel], completed work from [Notion database], merged pull requests from [GitHub repository], and continuity from yesterday's meeting notes.

Use these sections: CI speed; decisions; progress and changes; bugs and feedback; open questions and risks. Summarize only what changed, link important sources, and note any source that could not be read.
Step05

Notify the Team

End the agent instructions with a short Slack notification. Include the meeting date and page link, then verify that the message points to today's populated page rather than the template or yesterday's notes.

Example prompt
Post a brief, conversational message in [Slack channel] with the title and URL of today's completed standup pre-read. Mention the most important discussion item in one sentence.

What good looks like

  • Exactly one pre-read page is created for the current meeting.
  • The page covers the prior 24 hours and links important source items.
  • CI metrics match the configured Honeycomb query.
  • The team receives a brief Slack message with the correct page link.

Build your next product with ChatPRD

Turn an idea into a PRD, user stories, and a plan.

Try ChatPRD free

After the steps

Runbook notes

How to recover when the loop fails and where human judgment helps.

Recover

If it goes sideways

The agent writes to the wrong page or creates duplicates
Run the page-creation automation first, then identify the target by meeting date and template. Tell the agent to stop if it finds more than one matching page and report the duplicates.
The pre-read includes stale or unrelated activity
Set the team channel, project database, repository, time zone, and exact 24-hour window in the instructions. Require source links for decisions, bugs, and merged pull requests.
The Honeycomb metric is missing or incorrect
Test the Honeycomb MCP query by itself and save the working query in the agent instructions. Include the expected unit and comparison window.
The pre-read is too long to use in the meeting
Limit each section to the items that changed in the last 24 hours, combine duplicates, and move supporting links beneath a short summary.

Start shipping
better products.

Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.

Free to start
No credit card
SOC 2 certified
Enterprise ready