# Slack channel to searchable board

**What it does:** turns a Slack channel where your team shares links (inspiration, competitors, articles) into one web page of cards you can search and filter by tag.

**What it saved:** built in one night from 90 days of a real channel: 70 posts became 70 cards. Finding "that one site someone posted in May" goes from scrolling Slack to one click on a tag.

**Shown at:** AI Open House Session 003, 2026-09-25, by Colton Mulligan.

## What you need

| Tool | Cost |
|------|------|
| Claude (Claude Code or the Claude desktop app) | your existing plan |
| The Slack connector for Claude, with access to the channel | free |
| Somewhere to host one HTML file (Cloudflare Pages, Netlify, GitHub Pages) | $0 |

## How it works

1. Someone on the team posts a link in the channel with a comment, like "love the cursor effect."
2. Then Claude reads the channel through the Slack connector.
3. Then it writes one record per post: who posted, their words exactly as typed, every reply, and tags pulled from their words.
4. Then it builds one self-contained web page from those records.
5. Then you host that page anywhere static, marked noindex so search engines skip it.
6. Then anyone clicks a tag and the board filters to every post sharing it.

## The rules that matter

- Quote people word for word. If someone posted with no comment, the card says "no comment." Never let the AI invent a reason someone shared something.
- Tags come from the poster's own words. Tags inferred from the medium (video, font, site) are shown in lighter type so nobody mistakes them for a person's opinion.
- Slack's API only returns about 90 days of history on some plans. A workspace export (Slack admin, Settings, Import/Export Data) gives you the full archive as JSON to backfill from.
- If you schedule a daily refresh, add an alert for when it stops adding posts. Ours stopped quietly.

## The prompt

```text
Connect to my Slack and read the channel #[CHANNEL NAME] as far back as you can.

For every post that shares a link, image or file, make one record with:
- who posted it and the date
- exactly what they typed, word for word (if they wrote nothing, say "no comment", never invent a reason)
- every reply, with who wrote it
- 2 to 5 tags taken from the poster's own words
- a link back to the original Slack message

Then build a single web page from those records: a grid of cards, newest first,
a search box, and clickable tags that filter the board to everything sharing that tag.
Keep it one self-contained HTML file with no outside scripts.
Tell me how many posts you found and the oldest date you could reach.
```

## Stack card

```text
WHAT IT DOES:     Turns a Slack channel of shared links into a searchable, tag-filtered board
AI / MODEL:       Claude (Claude Code) with the Slack connector
STACK:            One JSON data file, one build script, static hosting
WHERE IT LIVES:   this recipe
COST TO RUN:      $0 a month to host
TIME SAVED:       One night to build; finding a past reference goes from scrolling to one click
TIER SHARED:      recipe
```

Built by Colton: https://builtbycolton.com

From the AI Open House free library: https://aiopenhouse.co/guests/#library