12 action items from 4 meetings today
4 recorded meetings produced 12 action items. 3 have no clear owner, all from the pricing review.
Every weekday at 17:30: the action items and owners from today's recorded Zoom meetings, from AI Companion summaries or transcripts.
What lands on your board
4 recorded meetings produced 12 action items. 3 have no clear owner, all from the pricing review.
A day of Zoom calls leaves a trail of "I'll send that over" that nobody wrote down. AI Companion writes a summary per meeting, but it lands in a separate email or tab for each one, and the items without a clear owner are exactly the ones that get lost.
At the end of each weekday the agent collects today's meeting summaries, or reads the transcript where there is no summary, and lists every action item with its owner, due date and meeting. Product managers and sales leads see in one place what they promised, and which items nobody took.
With an account-level app, call GET /v2/meetings/meeting_summaries?from=<today>&to=<today>. With a user-level app, list your recorded meetings with GET /v2/users/me/recordings?from=<today>&to=<today>.
Call GET /v2/meetings/{meetingUUID}/meeting_summary and read summary_overview, summary_details and next_steps. Double-encode a UUID that starts with / or contains //.
For a recorded meeting with no summary, take the file with file_type: TRANSCRIPT (VTT) from recording_files, download it from download_url with the access token, and pull out commitments: who said they would do what, and by when.
One row per action item with owner, due date if one was said, and the meeting. Items with no clear owner go first and are marked, and similar items from different meetings are merged.
With BRIEFKIN_TOKEN set, the agent sends the result as one card that updates in place every run. Without it, you get the report in the chat.
Pick your agent, where the data comes from and how to schedule it; the commands below follow.
One command. It puts the SKILL.md in ~/.claude/skills and in ~/.agents/skills, which Codex, Cursor, Gemini CLI, Copilot CLI and OpenCode read too.
shellcurl -fsSL https://briefkin.dev/skills/zoom-meeting-action-items/install | shAsk Claude Code:
promptRun the zoom-meeting-action-items skillOr paste the prompt instead of installing:
promptUsing the Zoom API, find my meetings from today that have an AI Companion meeting summary (GET /meetings/{meetingUUID}/meeting_summary) or a cloud recording transcript. From the summaries' next steps, or the transcripts where there is no summary, list every action item with its owner, the due date if one was said, and the meeting. Put items with no clear owner first, merge duplicates across meetings, and end with the one item most likely to be dropped.Weekdays at 17:30, from cron on any machine where Claude Code is signed in:
crontab30 17 * * 1-5 claude -p "Run the zoom-meeting-action-items skill"Add it with crontab -e. A laptop asleep at that time runs it when it wakes if you use launchd or Task Scheduler instead; GitHub Actions works for the API route.
Connect Briefkin once with curl -fsSL https://briefkin.dev/install | sh and every run lands on your board as a report card. How setup works
SKILL.md---
name: zoom-meeting-action-items
description: "Every weekday at 17:30: the action items and owners from today's recorded Zoom meetings, from AI Companion summaries or transcripts. Use when the user asks for this, or when it runs on a schedule."
---
# Zoom meeting action items digest
Every weekday at 17:30: the action items and owners from today's recorded Zoom meetings, from AI Companion summaries or transcripts.
A day of Zoom calls leaves a trail of "I'll send that over" that nobody wrote down. AI Companion writes a summary per meeting, but it lands in a separate email or tab for each one, and the items without a clear owner are exactly the ones that get lost.
At the end of each weekday the agent collects today's meeting summaries, or reads the transcript where there is no summary, and lists every action item with its owner, due date and meeting. Product managers and sales leads see in one place what they promised, and which items nobody took.
## Two ways to get the data
Use the API or MCP route when you have an MCP server, a CLI or an API token for the app. Otherwise, if you can drive a browser the user is signed in to (Claude in Chrome, a Playwright or Chrome DevTools MCP server), use the browser route. Say which one you used.
## Route A: API or MCP server
### What you need
- A Zoom app: a Server-to-Server OAuth app with `meeting:read:list_summaries:admin` and `meeting:read:summary:admin` for the whole account, or a user-level app with `meeting:read:summary` and `cloud_recording:read:list_user_recordings` for your own meetings. Your Zoom admin may need to approve it.
- AI Companion meeting summaries switched on (paid plans), or cloud recording with audio transcripts.
If something is missing, say what and stop; do not invent numbers.
### Steps
1. **Find today's meetings** (Zoom): With an account-level app, call `GET /v2/meetings/meeting_summaries?from=<today>&to=<today>`. With a user-level app, list your recorded meetings with `GET /v2/users/me/recordings?from=<today>&to=<today>`.
2. **Read each summary** (Zoom): Call `GET /v2/meetings/{meetingUUID}/meeting_summary` and read `summary_overview`, `summary_details` and `next_steps`. Double-encode a UUID that starts with `/` or contains `//`.
3. **Fall back to the transcript** (Zoom): For a recorded meeting with no summary, take the file with `file_type: TRANSCRIPT` (VTT) from `recording_files`, download it from `download_url` with the access token, and pull out commitments: who said they would do what, and by when.
4. **Write the digest**: One row per action item with owner, due date if one was said, and the meeting. Items with no clear owner go first and are marked, and similar items from different meetings are merged.
## Route B: the browser
### What you need
- Signed in to the Zoom web portal as the host of the meetings, or as an admin whose account-level AI Companion settings let them view summaries; summaries only exist for meetings where AI Companion was on.
If something is missing, say what and stop; do not invent numbers.
### Steps
Start at https://zoom.us/ai. The user is signed in already; never type passwords or codes, and ask if a sign-in page appears.
1. **Open your summaries**: Sign in to the Zoom web portal and open **AI Companion** in the left menu, then **Summaries** and **My Summaries** (also reachable from `https://zoom.us/ai`). Set the date filter to today.
2. **Read each summary**: Open each of today's meeting summaries and read the overview, the detailed sections and **Next steps**, which lists the action items with the names AI Companion assigned.
3. **Fall back to transcripts**: For a meeting with no summary, open **Recordings & Transcripts**, then **Cloud Recordings**, open today's recording and read the **Audio transcript** panel for commitments: who said they would do what, and by when.
4. **Write the digest**: One row per action item with owner, due date if one was said, and the meeting. Items with no clear owner go first and are marked; merge duplicates across meetings.
Note: Summaries appear a few minutes after a meeting ends, so run after the last meeting of the day. Only the host and people the host shared with see a summary in the portal, which is narrower than the account-level API. Transcripts are long; a day of recorded meetings is slow to read in a browser.
## The report
- **Action items** from today's meetings, with owner, due date and meeting.
- **Without an owner**: items nobody took, at the top.
- **Per meeting**: a count, and whether it came from a summary or a transcript.
- One line on the item most likely to be dropped.
## Post it to Briefkin
When `BRIEFKIN_TOKEN` is set, or `~/.config/briefkin/env` exists (the installer from https://briefkin.dev/docs writes it), also send the result to the user's Briefkin board as one card. Keep the `item_key` the same on every run so the card updates in place, use a new `event_id` per run, and put today's real numbers in place of the example ones. If the request fails, say so once and carry on.
```bash
[ -n "${BRIEFKIN_TOKEN:-}" ] || . ~/.config/briefkin/env
curl -sS -X POST "${BRIEFKIN_URL:-https://briefkin.dev}/v1/events" \
-H "authorization: Bearer $BRIEFKIN_TOKEN" \
-H "content-type: application/json" \
-d @- <<'JSON'
{
"v": 1,
"event_id": "zoom-meeting-action-items/<date of this run, e.g. 2026-10-16>",
"item_key": "zoom-meeting-action-items",
"project": "Acme",
"source": "Zoom meeting action items digest skill",
"kind": "update",
"title": "12 action items from 4 meetings today",
"summary": "4 recorded meetings produced **12 action items**. **3** have no clear owner, all from the pricing review.",
"widget": [
{
"type": "Table",
"props": {
"columns": [
"Action item",
"Owner",
"Due",
"Meeting"
],
"align": [
"left",
"left",
"left",
"left"
],
"rows": [
[
"Decide the cap on annual discounts",
"No owner",
"",
"Pricing review"
],
[
"Check how other tools price seats",
"No owner",
"",
"Pricing review"
],
[
"Draft the FAQ for plan changes",
"No owner",
"",
"Pricing review"
],
[
"Send revised pricing tiers to finance",
"Dana",
"Oct 20",
"Pricing review"
],
[
"Share churn interview notes",
"Lena",
"Oct 21",
"Customer research sync"
],
[
"Book two more interviews with admins",
"Lena",
"Oct 24",
"Customer research sync"
],
[
"Update the onboarding deck",
"Priya",
"",
"Customer research sync"
],
[
"Fix the flaky checkout test",
"Sam",
"Oct 20",
"Engineering standup"
],
[
"Review the SSO pull request",
"Theo",
"Oct 18",
"Engineering standup"
],
[
"Post standup notes in the team channel",
"Sam",
"Oct 17",
"Engineering standup"
],
[
"Send Contoso the security questionnaire",
"Marco",
"Oct 20",
"Contoso renewal call"
],
[
"Confirm the renewal start date with legal",
"Marco",
"Oct 22",
"Contoso renewal call"
]
]
}
},
{
"type": "Callout",
"props": {
"text": "The pricing review ended with three open items and no owners. Assign them before Monday's follow-up.",
"tone": "warn"
}
}
]
}
JSON
```
Widgets come from the Briefkin catalog (https://briefkin.dev/v1/catalog). Without a token, just give the user the report.
## Run it on a schedule
Suggested: Weekdays at 17:30 (`30 17 * * 1-5`). From cron, with the agent signed in on that machine:
```bash
30 17 * * 1-5 claude -p "Run the zoom-meeting-action-items skill"
# Codex: codex exec, Gemini CLI: gemini -p, Copilot CLI: copilot -p, OpenCode: opencode run
```
Or use the agent's own scheduler (Claude Code routines, Codex automations, Claude or ChatGPT scheduled tasks): https://briefkin.dev/skills/zoom-meeting-action-items#schedule
Source: https://briefkin.dev/skills/zoom-meeting-action-items
A Zoom app: a Server-to-Server OAuth app with meeting:read:list_summaries:admin and meeting:read:summary:admin for the whole account, or a user-level app with meeting:read:summary and cloud_recording:read:list_user_recordings for your own meetings. Your Zoom admin may need to approve it.
AI Companion meeting summaries switched on (paid plans), or cloud recording with audio transcripts.
https://mcp.zoom.us/mcp/zoom/streamable). It searches meetings, chat and docs, and fetches AI Companion summaries, transcripts, next steps, recordings and playback links. It needs a licensed user on a paid Zoom Workplace plan, works best with Smart Recording and Meeting Summary turned on, and sees only meetings you hosted or were given access to. Connectors added in claude.ai also appear in Claude Code when you sign in with that account (docs).claude mcp add takes them with --client-id and --client-secret (docs).ZOOM_ACCOUNT_ID, ZOOM_CLIENT_ID and ZOOM_CLIENT_SECRET; the skill exchanges them for a token at https://zoom.us/oauth/token with grant_type=account_credentials.A Server-to-Server app reads across the whole account, so give it read scopes only and nothing it doesn't report on.
AI Companion meeting summaries need a paid Zoom plan and have to be switched on for the meeting. Without them the skill reads cloud recording transcripts, which need a Pro plan or higher with cloud recording and audio transcripts on. A meeting that was neither summarized nor recorded leaves nothing to read.
Listing summaries across the account uses an admin scope, which in practice means a Server-to-Server OAuth app set up by a Zoom admin. With an app for your own user, the skill lists your recordings and fetches each meeting's summary by ID, which covers the meetings you hosted.
No. The skill runs with your own agent and gives you the report in the chat. With a Briefkin token set, each run also posts it to your board as a report card, so you can glance at it instead of asking again.
Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot CLI, OpenCode, Windsurf and Amp all read SKILL.md files. The install command puts it in ~/.claude/skills and ~/.agents/skills, where they look. ChatGPT, the Claude app and Grok take the prompt instead. Pick your agent under "Install and run it" and the commands change to match.
Yes. Switch the data source to "Your browser" and the steps change to the Zoom pages themselves: the agent opens zoom.us/ai in a browser you are signed in to and reads the numbers off the page. Claude Code, Codex, Cursor, Claude app and ChatGPT can drive a browser out of the box; Gemini CLI, Copilot CLI and OpenCode need a Playwright or Chrome DevTools MCP server first. It is slower than the API and needs you signed in, so it suits a report you run by hand or from a scheduler on your own computer.
Skills are free to use with any agent. Briefkin is where their results land: one board for every report, number and question your agents send, with a ping when something needs you.
Free during the open beta. No card.