4 threads waiting on your reply, oldest 6 days
Want me to draft replies? Drafts go into each thread and nothing is sent.
4 people are waiting on you. Hannah Cole asked about pilot pricing 6 days ago.
Weekday afternoons: Gmail threads where someone else wrote last over two days ago, ranked by how long they have waited, with an offer to draft replies.
Reads
You get
The card on the right is a sample. Yours shows your own numbers and updates in place every run.
What lands on your board
Want me to draft replies? Drafts go into each thread and nothing is sent.
4 people are waiting on you. Hannah Cole asked about pilot pricing 6 days ago.
The emails that hurt are not the ones you decided to ignore. They are the ones you read on your phone, meant to answer properly, and then lost below the fold. Two days later the sender has followed up, or quietly given up.
This skill looks for exactly that shape: a thread where the last message is from someone else, a real person, and older than 48 hours. It ranks them by how long they have waited and offers to draft replies into each thread for you to edit and send.
Call users.threads.list with q: "in:inbox older_than:2d newer_than:30d -category:promotions -category:social -category:forums -category:updates". A thread matches if any of its messages fits, so the real test is the next step.
For each thread, users.threads.get with format: metadata and take the last message. Keep the thread if its From is not you (your address from users.getProfile, aliases from users.settings.sendAs.list) and its internalDate is more than 48 hours old.
Skip messages with a List-Unsubscribe header, senders like noreply@ or notifications@, calendar invitations, and threads where you are only in Cc.
Order by hours waiting, read the last message to state the ask in one line, and ask which ones to draft. Drafts go into the thread with users.drafts.create and the thread's threadId; nothing is sent.
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/gmail-emails-waiting-on-reply/install | shAsk Claude Code:
promptRun the gmail-emails-waiting-on-reply skillOr paste the prompt instead of installing:
promptFind Gmail threads in my inbox from the last 30 days where the last message is from someone other than me (check my address and send-as aliases) and is more than 48 hours old. Skip newsletters and anything with a List-Unsubscribe header, noreply senders, calendar invitations and threads where I'm only in Cc. Rank them by how long they have waited and give me the sender, subject, the ask in one line and days waiting. Then ask me which ones to draft replies for. Create drafts inside the original threads only after I answer, and never send anything.Weekdays at 16:00, from cron on any machine where Claude Code is signed in:
crontab0 16 * * 1-5 claude -p "Run the gmail-emails-waiting-on-reply 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 question for you. How setup works
SKILL.md---
name: gmail-emails-waiting-on-reply
description: "Weekday afternoons: Gmail threads where someone else wrote last over two days ago, ranked by how long they have waited, with an offer to draft replies. Use when the user asks for this, or when it runs on a schedule."
---
# Gmail threads waiting on your reply
Weekday afternoons: Gmail threads where someone else wrote last over two days ago, ranked by how long they have waited, with an offer to draft replies.
The emails that hurt are not the ones you decided to ignore. They are the ones you read on your phone, meant to answer properly, and then lost below the fold. Two days later the sender has followed up, or quietly given up.
This skill looks for exactly that shape: a thread where the last message is from someone else, a real person, and older than 48 hours. It ranks them by how long they have waited and offers to draft replies into each thread for you to edit and send.
## 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 Gmail connector or MCP server, or a Google OAuth client with `gmail.readonly`, to find the threads.
- The `gmail.compose` scope as well if the agent should write drafts into them.
If something is missing, say what and stop; do not invent numbers.
### Steps
1. **Search for candidates** (Gmail): Call `users.threads.list` with `q: "in:inbox older_than:2d newer_than:30d -category:promotions -category:social -category:forums -category:updates"`. A thread matches if any of its messages fits, so the real test is the next step.
2. **Check who wrote last** (Gmail): For each thread, `users.threads.get` with `format: metadata` and take the last message. Keep the thread if its `From` is not you (your address from `users.getProfile`, aliases from `users.settings.sendAs.list`) and its `internalDate` is more than 48 hours old.
3. **Drop automated mail**: Skip messages with a `List-Unsubscribe` header, senders like `noreply@` or `notifications@`, calendar invitations, and threads where you are only in Cc.
4. **Rank and ask** (Gmail): Order by hours waiting, read the last message to state the ask in one line, and ask which ones to draft. Drafts go into the thread with `users.drafts.create` and the thread's `threadId`; nothing is sent.
## Route B: the browser
### What you need
- Signed in to Gmail in the browser, with Drafts allowed (it is by default). The person checks each draft in Drafts before sending.
If something is missing, say what and stop; do not invent numbers.
### Steps
Start at https://mail.google.com/. The user is signed in already; never type passwords or codes, and ask if a sign-in page appears.
1. **Search the candidates**: In the search box type `in:inbox older_than:2d newer_than:30d -category:promotions -category:social -category:forums -category:updates` and press Enter. The list shows the thread's last sender and the date, which is enough to skip most threads.
2. **Check who wrote last**: Open each thread where the last sender shown is not you and the date is more than two days old. Confirm the final message is from someone else and that you are in **To**, not Cc (open **Show details** under the sender). Your own aliases are listed under **Settings > See all settings > Accounts > Send mail as**.
3. **Drop the automated ones**: Skip anything with an **Unsubscribe** link by the sender name, `noreply` or `notifications` addresses, and calendar invitations.
4. **Rank and offer drafts**: Read the last message and write the ask in one line, order by days waiting and ask which to answer. On a yes, click **Reply** inside the thread, write the draft and click away; Gmail saves it to Drafts in that thread. Never click **Send**.
Note: The search returns threads where any message matches the dates, so the agent has to open borderline threads to be sure. Drafts written in the browser land in the thread the same way the API route's do; check them under Drafts before sending.
## The report
- How many threads are waiting on you, and the age of the oldest.
- Per thread: sender, subject, the ask in one line, days waiting.
- The question: draft replies for all, some or none.
## 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": "gmail-emails-waiting-on-reply/<date of this run, e.g. 2026-10-16>",
"item_key": "gmail-emails-waiting-on-reply",
"project": "Inbox",
"source": "Gmail threads waiting on your reply skill",
"kind": "action",
"title": "4 threads waiting on your reply, oldest 6 days",
"summary": "**4** people are waiting on you. Hannah Cole asked about pilot pricing **6 days** ago.",
"action": {
"request": "Want me to draft replies? Drafts go into each thread and nothing is sent.",
"options": [
"Draft all 4",
"Draft the 2 oldest",
"Remind me tomorrow",
"They are handled"
]
},
"widget": {
"type": "Table",
"props": {
"columns": [
"From",
"Subject",
"Waiting"
],
"align": [
"left",
"left",
"right"
],
"rows": [
[
"Hannah Cole",
"Pilot pricing for 40 seats",
"6 days"
],
[
"Marco Bianchi",
"Podcast interview, November dates",
"4 days"
],
[
"Aisha Karim",
"Reference check for J. Park",
"3 days"
],
[
"Tom Reyes",
"Can we move our 1:1?",
"2 days"
]
]
}
}
}
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 16:00 (`0 16 * * 1-5`). From cron, with the agent signed in on that machine:
```bash
0 16 * * 1-5 claude -p "Run the gmail-emails-waiting-on-reply 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/gmail-emails-waiting-on-reply#schedule
Source: https://briefkin.dev/skills/gmail-emails-waiting-on-reply
A Gmail connector or MCP server, or a Google OAuth client with gmail.readonly, to find the threads.
The gmail.compose scope as well if the agent should write drafts into them.
https://gmailmcp.googleapis.com/mcp/v1, in developer preview with the other Google Workspace MCP servers. It needs an OAuth client in Google Cloud set up for an AI agent (setup).gmail.readonly scope, and store GMAIL_CLIENT_ID, GMAIL_CLIENT_SECRET and GMAIL_REFRESH_TOKEN. Publish the OAuth app (or make it Internal in Workspace): refresh tokens from an app left in Testing expire after seven days.gmail.readonly can search and read but not send, delete or relabel; add gmail.compose only if you want the agent to leave drafts.
Yes. Sent mail is part of the thread whatever device it came from. A reply from a different account is missed unless that address is set up as a send-as alias in Gmail settings.
Under two days, most threads are just today's mail. Change the search (older_than:1d, older_than:5d) and the 48-hour check together, for example one day for customer email and a week for everything else.
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 asks you, 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 Gmail pages themselves: the agent opens mail.google.com 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.