惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

月光博客
月光博客
D
Docker
腾讯CDC
J
Java Code Geeks
大猫的无限游戏
大猫的无限游戏
The Cloudflare Blog
Martin Fowler
Martin Fowler
MongoDB | Blog
MongoDB | Blog
博客园 - Franky
博客园 - 三生石上(FineUI控件)
Recent Announcements
Recent Announcements
F
Fortinet All Blogs
IT之家
IT之家
WordPress大学
WordPress大学
M
MIT News - Artificial intelligence
爱范儿
爱范儿
Microsoft Azure Blog
Microsoft Azure Blog
Vercel News
Vercel News
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
小众软件
小众软件
N
Netflix TechBlog - Medium
T
Tailwind CSS Blog
Engineering at Meta
Engineering at Meta
博客园 - 【当耐特】

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
We moved daily standups into Slack
Kelly Lewand · 2026-05-14 · via DEV Community

Every standup tool I've shipped, used, or watched a team adopt eventually loses to the same thing: people forget to open it.

The product is fine. The form works. The prompts are reasonable. But it sits behind a tab that nobody has open at 9:14am, and so the submissions trickle in at 10:30, or 11, or never. By Wednesday the manager is back to chasing status in DMs. By Friday the standup tool is a tab the team closes on Monday morning and reopens when someone reminds them.

We hit this on our own product. So we moved the prompt into Slack.

The thing that's actually broken

Standups don't break because the questions are bad. They break because of friction, which mostly shows up in three places:

  • Context-switch cost. Your engineer is in a PR, an IDE, a Linear ticket, and Slack. The standup tool is a fifth tab. The tax to switch and type the same thing in a slightly different shape is what gets skipped.
  • The 9am punctuality assumption. Standup tools assume people will open them at the standup time. People don't. They open them when they remember, which is rarely when you'd like.
  • Manager chasing. When submissions are sparse, the manager opens DMs and asks individuals. The standup tool now exists to be ignored, while the actual standup happens one-to-one in DMs the rest of the team can't see.

The fix everyone tries first is a reminder. Email at 9:01, push notification at 9:05, Slackbot at 9:10. Reminders raise the submission rate a little. They don't fix the underlying thing, which is that the prompt is in the wrong place.

standup no one opens

What changes when the prompt lives in Slack

We rebuilt the loop so the bot DMs you the standup questions at the configured time, and you reply in the thread. Plain English. However you'd answer if someone asked you in person.

No form, no buttons. You type "shipped the auth migration last night, today wrapping the rollout, blocked on the design review for the new settings page" and the AI splits that into the right answers under each question.

The DM is the prompt. There's no separate tool to open. The notification you already see, in the app you're already in, is the prompt itself. The 9am punctuality problem disappears because the notification is the standup.

Late submissions don't pollute the channel. When someone submits an hour after the window closes, the late reply appends into the existing thread instead of creating a second top-level post. The channel scroll stays clean. The summary still updates.

The summary lives where the team reads. Once the standup window closes, the bot posts a single tidy summary to your team channel: AI-generated TL;DR up top, who's blocked, who's asking for help, with each person's submission threaded underneath. Stakeholders who don't want to read every reply can read the top of the post. Engineers who want the detail open the thread.

I work on Kollabe, so take the workflow with the appropriate grain of salt. The pattern generalises. If your tool of choice can prompt in Slack, parse a plain-English reply, and post a channel summary, you can do the same thing. The point isn't which tool. The point is where the prompt lives.

Where the opposing view is right

Dedicated standup tools aren't pointless. There are real cases:

  • Very small teams with one engineer and one PM don't need any of this. A Slack channel and the muscle of habit is fine.
  • Heavily regulated environments where every submission needs an audit trail with strict identity assertions sometimes need the tool to be the source of record, not Slack.
  • Async-only teams across many timezones sometimes prefer the long-form, properly-threaded view a standup tool provides for retros and recaps.

The honest version of the argument is: if your standups already work, don't touch them. If your standups exist but everyone ignores the tool and you're DM-chasing by Wednesday, the prompt is in the wrong place.

The channel summary

A few patterns we've watched teams use

  • Standups without the meeting. Engineers reply async over morning coffee. Managers scan the channel summary for blockers around lunch. No 9am call, no roll-call awkwardness, the blockers still get seen.
  • Distributed teams with no shared morning. The DM lands at each team's configured time. Nobody gets pinged at 6am because someone in another timezone forgot to think about it.
  • Cross-team visibility for PMs and design. Drop the daily summary into a wider #standups-engineering channel that PMs and design lurk in. They get context without being invited to a meeting.
  • Hybrid live-and-async. Pre-submitted answers via Slack, then a 15-minute call where the time is spent on the blockers and help-asks in the summary, not on roll-call.

One more thing: Claude reading across standups

Once the standup record exists somewhere structured, you can do a lot with it. Every Kollabe standup is reachable through MCP, which means you can wire it to Claude (or any MCP-capable client) and ask things the Slack channel can't answer on its own.

Things I personally now ask Claude on Monday morning:

  • "What did the platform team ship last week?"
  • "Has anyone mentioned the data migration in the last 14 days? Who's still blocked?"
  • "Grab this month's standup summaries and give me a release-notes draft."
  • "What's everyone working on today across all my teams?"

The Slack pattern handles the daily pulse. The MCP layer handles zooming out: weekly recaps, monthly retros, patterns across teams you don't sit on. They cover different jobs. Standups in Slack lets the team write less. Claude over the standup record lets the manager read less. Both are wins.

If you want the MCP setup specifically, we have a guide for Claude Desktop. It's a ten-minute job.

claude-standup

The rule of thumb

Before you pick a standup tool, ask one question: is the prompt going to land in a place my team already opens?

If the answer is yes, the rest is implementation detail. If the answer is no, if the tool requires a habit your team doesn't have, you're going to spend the next six months trying to install that habit. You won't win. The tool will lose to the tab people already have open.

For us that place is Slack. For you it might be Teams. For a hardware team it might be a physical board in a hallway. The medium doesn't matter. What matters is that the prompt meets the team where they already are.

Closing

Microsoft Teams is next on our list, with the same shape: conversational DM submission, AI-parsed answers, a channel summary at the end of the window. Same idea, different chat app.

If you want to try the Slack version, we've shipped it on Kollabe Premium and Enterprise. Your team installs Slack at the org level, you flip on DM reminders for the standup, pick a channel for the summary, choose what time it should land. About four clicks total.

But honestly, even if you don't use Kollabe, the takeaway you can use tomorrow is this: stop trying to make people open a separate tool for a 30-second update. Move the prompt into the place they already are. The submission rate problem solves itself.