What it does
The CSAT percentage is the number leadership asks for. The comments are the part a support lead actually learns from, and Zendesk keeps them a few clicks deep in the satisfaction report, one ticket at a time.
Every Monday this skill pulls last week's ratings, computes the score the way Zendesk does, and puts every bad rating's comment on one page with its ticket number and agent. Patterns show up quickly when the comments sit side by side: three customers mentioning a slow refund is a process problem, not three unlucky tickets.
How it works
1Pull last week's ratings
Call
GET /api/v2/satisfaction_ratings?score=received&start_time=<last Monday 00:00, Unix>&end_time=<Sunday 23:59:59>and follow the pagination links. Keepscore,comment,reason,ticket_id,assignee_idandgroup_id.- 2
2Compute the score
CSAT is good divided by good plus bad. For the response rate, make the same call with
score=offered(surveys sent but not answered) and divide received by received plus offered. 3Look up the bad ones
For every
badrating, get the ticket subject withGET /api/v2/tickets/show_many?ids=..., the agent name withGET /api/v2/users/show_many?ids=..., and the first reply time fromGET /api/v2/tickets/{ticket_id}/metrics(reply_time_in_minutes).- 4
4Group the comments
Sort the bad ratings into a few plain themes from their comments and reasons (slow reply, not resolved, policy, product bug) and count each. Quote comments as written, minus names, emails and order numbers.
- 5
5Post the score
Headline: CSAT for the week against the week before. Add ratings received, the response rate, the themes with counts and the quoted comments.
Post it to the board
With
BRIEFKIN_TOKENset, the agent sends the result as one card that updates in place every run. Without it, you get the report in the chat.
Install and run it
Pick your agent, where the data comes from and how to schedule it; the commands below follow.
Install the skill
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.
shell
curl -fsSL https://briefkin.dev/skills/zendesk-weekly-csat/install | shRun it
Ask Claude Code:
prompt
Run the zendesk-weekly-csat skillOr paste the prompt instead of installing:
prompt
Using the Zendesk Satisfaction Ratings API, get every rating received last week (Monday to Sunday). Compute CSAT as good / (good + bad), and the response rate using the surveys that were offered but not answered. For every bad rating, give me the ticket number, assignee, reason and comment as written, and the ticket's first reply time. Group the bad comments into a few themes with counts. Compare CSAT with the week before if you can, and end with the one theme I should bring to the team meeting.Put it on a schedule
Mondays at 9:30, from cron on any machine where Claude Code is signed in:
crontab
30 9 * * 1 claude -p "Run the zendesk-weekly-csat 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.Send it to your board
Connect Briefkin once with
curl -fsSL https://briefkin.dev/install | shand every run lands on your board as a stat tile. How setup works
View the SKILL.mdRaw file
SKILL.md---
name: zendesk-weekly-csat
description: "Monday mornings: last week's Zendesk satisfaction score, how many customers rated, and every comment that came with a bad rating. Use when the user asks for this, or when it runs on a schedule."
---
# Zendesk weekly CSAT report
Monday mornings: last week's Zendesk satisfaction score, how many customers rated, and every comment that came with a bad rating.
The CSAT percentage is the number leadership asks for. The comments are the part a support lead actually learns from, and Zendesk keeps them a few clicks deep in the satisfaction report, one ticket at a time.
Every Monday this skill pulls last week's ratings, computes the score the way Zendesk does, and puts every bad rating's comment on one page with its ticket number and agent. Patterns show up quickly when the comments sit side by side: three customers mentioning a slow refund is a process problem, not three unlucky tickets.
## 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 Zendesk admin login (the Satisfaction Ratings API is admin only) with an API token in `ZENDESK_EMAIL` and `ZENDESK_API_TOKEN`, or a Zendesk connector with the same rights.
- CSAT surveys switched on for solved tickets.
If something is missing, say what and stop; do not invent numbers.
### Steps
1. **Pull last week's ratings** (Zendesk): Call `GET /api/v2/satisfaction_ratings?score=received&start_time=<last Monday 00:00, Unix>&end_time=<Sunday 23:59:59>` and follow the pagination links. Keep `score`, `comment`, `reason`, `ticket_id`, `assignee_id` and `group_id`.
2. **Compute the score**: CSAT is good divided by good plus bad. For the response rate, make the same call with `score=offered` (surveys sent but not answered) and divide received by received plus offered.
3. **Look up the bad ones** (Zendesk): For every `bad` rating, get the ticket subject with `GET /api/v2/tickets/show_many?ids=...`, the agent name with `GET /api/v2/users/show_many?ids=...`, and the first reply time from `GET /api/v2/tickets/{ticket_id}/metrics` (`reply_time_in_minutes`).
4. **Group the comments**: Sort the bad ratings into a few plain themes from their comments and reasons (slow reply, not resolved, policy, product bug) and count each. Quote comments as written, minus names, emails and order numbers.
5. **Post the score**: Headline: CSAT for the week against the week before. Add ratings received, the response rate, the themes with counts and the quoted comments.
## Route B: the browser
### What you need
- Signed in to Zendesk as an agent with Explore Viewer access; the Rated tickets view is a default view every agent can open.
If something is missing, say what and stop; do not invent numbers.
### Steps
Start at https://www.zendesk.com/login/. The user is signed in already; never type passwords or codes, and ask if a sign-in page appears.
1. **Open the Satisfaction tab**: Enter your subdomain on the sign-in page, open **Explore** from the product tray, then **Dashboards** > **Zendesk Support** > **Satisfaction**. Set the time filter to **Last week** (Monday to Sunday) and read the satisfaction score and the number of rated tickets; change the filter to the week before for the comparison.
2. **Get the response rate**: Read the survey response rate on the same tab where your dashboard shows it; otherwise divide rated tickets by the tickets solved last week from the **Tickets** tab with the same time filter.
3. **Read the bad comments**: Switch to Support and open **Views** > **Rated tickets from the last 7 days**, sort by the **Satisfaction** column and open each ticket rated **Bad**. The rating comment and reason sit above the conversation; note the assignee and the gap between the first message and the first public agent reply.
4. **Group and post**: Sort the bad comments into a few plain themes (slow reply, not resolved, policy, product bug), quote them as written minus names and order numbers, and post the score, the rating count, the response rate, the themes and the comments.
Note: The Rated tickets view covers only the last 7 days, so run this on Monday morning. Newer customizable CSAT surveys with 1 to 5 scales are mapped to good and bad in Explore, so the score matches the API route. Explore refreshes about hourly; ratings that arrive on Monday morning may be missing from both.
## The report
- **CSAT** for last week (good divided by good plus bad) and the week before.
- **Ratings received** and the survey response rate.
- **Bad ratings by theme**, with counts.
- **The comments**, quoted, with ticket number, agent and first reply time.
- One line on the theme to bring to the team.
## 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": "zendesk-weekly-csat/<date of this run, e.g. 2026-10-16>",
"item_key": "zendesk-weekly-csat",
"project": "Acme",
"source": "Zendesk weekly CSAT report skill",
"kind": "metric",
"title": "CSAT 91% last week, 14 bad ratings",
"metric": {
"label": "CSAT",
"value": 91,
"unit": "%",
"better": "higher",
"period_start": "<start of the period, ISO 8601>",
"period_end": "<end of the period, ISO 8601>"
},
"glance": {
"type": "Grid",
"props": {
"columns": 3
},
"children": [
{
"type": "Stat",
"props": {
"label": "Ratings",
"value": 156
}
},
{
"type": "Stat",
"props": {
"label": "Bad ratings",
"value": 14,
"tone": "bad"
}
},
{
"type": "Stat",
"props": {
"label": "Response rate",
"value": 23,
"unit": "%"
}
}
]
},
"widget": [
{
"type": "Bars",
"props": {
"items": [
{
"label": "Slow reply",
"value": 5
},
{
"label": "Not resolved",
"value": 4
},
{
"label": "Refund policy",
"value": 3
},
{
"label": "Product bug",
"value": 2
}
],
"unit": "ratings",
"tone": "bad"
}
},
{
"type": "Table",
"props": {
"columns": [
"Ticket",
"Agent",
"Comment"
],
"rows": [
[
"#30411",
"Aisha",
"\"Waited two days for an answer about a double charge.\""
],
[
"#30397",
"Aisha",
"\"Had to explain the problem three times.\""
],
[
"#30388",
"Jonas",
"\"Ticket was closed but the export still fails.\""
],
[
"#30452",
"Marco",
"\"Ten business days for a refund is too long.\""
],
[
"#30470",
"Lena",
"\"This bug has been known for a month and is still not fixed.\""
]
]
}
},
{
"type": "Callout",
"props": {
"text": "Slow reply is behind 5 of the 14 bad ratings, and 4 of those tickets waited more than 24 hours for a first reply.",
"tone": "warn"
}
}
],
"size": "md"
}
JSON
```
Widgets come from the Briefkin catalog (https://briefkin.dev/v1/catalog). Without a token, just give the user the report.
Set `period_start` and `period_end` to the window the number covers: the tile names it from them ("Last 7 days", "Today", "This month"). For a count right now (open items, a balance), send `"period_label": ""` so no period shows; for another name, send it as `period_label` (at most 24 characters).
## Run it on a schedule
Suggested: Mondays at 9:30 (`30 9 * * 1`). From cron, with the agent signed in on that machine:
```bash
30 9 * * 1 claude -p "Run the zendesk-weekly-csat 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/zendesk-weekly-csat#schedule
Source: https://briefkin.dev/skills/zendesk-weekly-csat
What your agent needs
A Zendesk admin login (the Satisfaction Ratings API is admin only) with an API token in
ZENDESK_EMAILandZENDESK_API_TOKEN, or a Zendesk connector with the same rights.CSAT surveys switched on for solved tickets.
Connecting Zendesk
- MCP server. Zendesk doesn't list an official one; community MCP servers exist, and they need the same subdomain, email and API token as a direct API call.
- API with an API token. An admin enables token access in Admin Center under Apps and integrations > APIs and creates a token. The skill reads
ZENDESK_SUBDOMAIN,ZENDESK_EMAILandZENDESK_API_TOKENand sends basic auth as{email}/token:{token}. - OAuth client. For a narrower credential, register an OAuth client in Admin Center and request the
readscope.
An API token acts as the user who made it, with that agent's full rights; an OAuth token with only read can't update, assign or close tickets.
Questions
Why is the score good divided by good plus bad?
That is how Zendesk reports satisfaction: customers who did not answer are left out rather than counted as unhappy. If you use the newer customizable CSAT survey with a 1 to 3 or 1 to 5 scale, Zendesk maps the top score (3, or 4 and 5) to good and the rest to bad, so the same formula holds. The response rate next to it tells you how many customers the score speaks for.
A customer changed their rating after we followed up. Which one counts?
The current one. The API returns each rating with its latest score and comment, so a bad rating that a customer changed to good drops off the bad list on the next run.
Do I need Briefkin to use the Zendesk weekly CSAT report skill?
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 stat tile, so you can glance at it instead of asking again.
Which agents can run it?
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.
Can it work without an API or MCP server, from my browser?
Yes. Switch the data source to "Your browser" and the steps change to the Zendesk pages themselves: the agent opens www.zendesk.com/login 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.