Catch every Stripe dispute that needs evidence before its deadline

Every morning: Stripe disputes and inquiries waiting on a response, with amount, reason and the evidence due date, and a question on who gathers it.

Runs
Daily at 9:00
Uses
Stripe
Posts
Asks you

From your apps to your board

How the skill works: your agent reads open disputes from Stripe; who can gather evidence from What you tell it, daily at 9:00, and gives you a question for you and a telegram ping and the report in the chat.

Reads

  • StripeOpen disputes
  • What you tell itWho can gather evidence
reads
Your agentClaude Code, Codex, Cursor or GeminiDaily at 9:00
posts

You get

  • A question for youOn your board, answered with a tap
  • A Telegram pingOnly when it needs your answer, if you connect Telegram
  • The report in the chatEvery run, with or without Briefkin

What you get

  • Open disputes: how many and the amount at stake, with the earliest deadline.
  • The list, soonest first: customer, amount, reason, evidence due date and whether evidence is already saved.
  • What to collect for each reason.
  • The ask: who gathers the evidence, or which small one to accept.

The card on the right is a sample. Yours shows your own numbers and updates in place every run.

What lands on your board

AcmeStripe disputes to respond to · Claude Code

4 disputes need evidence, the first due in 3 days

Who gathers the evidence for these 4 disputes?

4 open disputes worth $1,386 need a response. The $612 fraud claim is due Oct 20 and has no evidence saved yet.

CustomerAmountReasonDueEvidence
j.martin@example.com$612.00FraudulentOct 20None
orders@northwind.example$480.00Product not receivedOct 23Tracking saved
li.chen@example.com$245.00Subscription canceledOct 27None
k.okafor@example.com$49.00Duplicate (inquiry)Oct 29None

The $49 duplicate is still an inquiry: answering it now can keep it from becoming a chargeback.

Resolve
An example with made-up numbers. Your agent posts your real ones.

What it does

A dispute you don't answer is a dispute you lose, and the deadline sits in a field most people only see when they open the Disputes page. Small teams miss them in the gap between "finance will handle it" and "support has the emails".

Every morning the agent lists the disputes and inquiries that still need a response, sorted by how soon evidence is due, with the reason the bank gave and what kind of evidence answers it. The card asks who takes each one, so the question gets an owner the same day it arrives.

How it works

  1. 1List the open disputes

    Call GET /v1/disputes with created[gte] 120 days ago and limit=100, paging with starting_after. The list has no status filter, so keep the disputes whose status is needs_response or warning_needs_response (an inquiry that is not yet a chargeback).

  2. 2Add the payment behind each one

    Retrieve the charge (GET /v1/charges/{id}) for the customer email, description and payment date. Note evidence_details.has_evidence (evidence already saved) and evidence_details.submission_count.

  3. 3

    3Rank by deadline

    Sort by evidence_details.due_by (a Unix timestamp), soonest first, and mark anything due within 72 hours as urgent. Match the reason to the evidence that answers it: tracking for product_not_received, the cancellation policy and logs for subscription_canceled, the second charge for duplicate, sign-in and delivery records for fraudulent.

  4. 4

    4Ask who gathers the evidence

    Post the list as an ask and record the answer in a local notes file next to the dispute ID. The skill writes nothing to Stripe: an update with evidence is sent to the bank unless submit=false is passed, so evidence stays a job for a person in the Dashboard.

  5. Post it to the board

    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.

Install and run it

Pick your agent, where the data comes from and how to schedule it; the commands below follow.

Agent
Data
Schedule
  1. 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.

    shellcurl -fsSL https://briefkin.dev/skills/stripe-disputes-to-respond/install | sh
  2. Run it

    Ask Claude Code:

    promptRun the stripe-disputes-to-respond skill

    Or paste the prompt instead of installing:

    promptUsing the Stripe API (GET /v1/disputes, created in the last 120 days, all pages), list the disputes whose status is needs_response or warning_needs_response. For each, give the customer email from the charge, the amount, the reason, evidence_details.due_by as a date, and whether evidence is already saved. Sort by due date, mark anything due within 72 hours, and say what evidence answers each reason. Then ask me who should gather the evidence for each one. Do not update or submit anything in Stripe.
  3. Put it on a schedule

    Daily at 9:00, from cron on any machine where Claude Code is signed in:

    crontab0 9 * * * claude -p "Run the stripe-disputes-to-respond 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.

  4. Send it to your board

    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

View the SKILL.mdRaw file
SKILL.md---
name: stripe-disputes-to-respond
description: "Every morning: Stripe disputes and inquiries waiting on a response, with amount, reason and the evidence due date, and a question on who gathers it. Use when the user asks for this, or when it runs on a schedule."
---

# Stripe disputes to respond to

Every morning: Stripe disputes and inquiries waiting on a response, with amount, reason and the evidence due date, and a question on who gathers it.

A dispute you don't answer is a dispute you lose, and the deadline sits in a field most people only see when they open the Disputes page. Small teams miss them in the gap between "finance will handle it" and "support has the emails".

Every morning the agent lists the disputes and inquiries that still need a response, sorted by how soon evidence is due, with the reason the bank gave and what kind of evidence answers it. The card asks who takes each one, so the question gets an owner the same day it arrives.

## 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

- Read access to disputes and charges: the Stripe MCP server (`https://mcp.stripe.com`) with OAuth or a restricted key as the bearer token, which reads them through its `list_disputes` and search tools, or a restricted key with Disputes and Charges set to Read in `STRIPE_API_KEY`.
- Who on the team can gather evidence (support, fulfillment, finance), so the card can offer them as answers.

If something is missing, say what and stop; do not invent numbers.

### Steps

1. **List the open disputes** (Stripe): Call `GET /v1/disputes` with `created[gte]` 120 days ago and `limit=100`, paging with `starting_after`. The list has no status filter, so keep the disputes whose `status` is `needs_response` or `warning_needs_response` (an inquiry that is not yet a chargeback).
2. **Add the payment behind each one** (Stripe): Retrieve the `charge` (`GET /v1/charges/{id}`) for the customer email, description and payment date. Note `evidence_details.has_evidence` (evidence already saved) and `evidence_details.submission_count`.
3. **Rank by deadline**: Sort by `evidence_details.due_by` (a Unix timestamp), soonest first, and mark anything due within 72 hours as urgent. Match the `reason` to the evidence that answers it: tracking for `product_not_received`, the cancellation policy and logs for `subscription_canceled`, the second charge for `duplicate`, sign-in and delivery records for `fraudulent`.
4. **Ask who gathers the evidence**: Post the list as an ask and record the answer in a local notes file next to the dispute ID. The skill writes nothing to Stripe: an update with evidence is sent to the bank unless `submit=false` is passed, so evidence stays a job for a person in the Dashboard.

## Route B: the browser

### What you need

- Signed in to the Stripe Dashboard with a role that can view disputes and payments (Administrator, Analyst, or a Support role); 2FA already done.

If something is missing, say what and stop; do not invent numbers.

### Steps

Start at https://dashboard.stripe.com/disputes. The user is signed in already; never type passwords or codes, and ask if a sign-in page appears.

1. **Open the disputes that need a response**: On the **Disputes** page in the Stripe Dashboard choose the **Needs response** filter. Make sure the Dashboard is in live mode, not the sandbox, and that the right account is selected in the top left if you have several. Inquiries appear in the same list, labeled as inquiries.
2. **Read each row**: For each dispute note the customer, the amount, the **Reason** and the **Respond by** date. Open the dispute to see the payment's date and description, and whether the evidence section already holds files (the page says so under the response status).
3. **Sort by deadline**: Order the list by Respond by, soonest first, and mark anything due within 72 hours. For each reason, write down what answers it: tracking for product not received, the cancellation policy and logs for subscription canceled, the second charge for duplicate, sign-in and delivery records for fraudulent.
4. **Stop before responding**: Do not click **Submit evidence** or **Accept dispute**; both are final. Post the list and the question of who gathers the evidence, and leave the response itself to the person in the Dashboard.

Note: The Dashboard already does most of this report: the Needs response list, the deadline and the evidence status are all on one page, so the browser route is a good fit here. The one thing to watch is the account picker: the list is per Stripe account, so a business with several accounts runs the steps once per account.

## The report

- **Open disputes**: how many and the amount at stake, with the earliest deadline.
- **The list**, soonest first: customer, amount, reason, evidence due date and whether evidence is already saved.
- **What to collect** for each reason.
- The ask: who gathers the evidence, or which small one to accept.

## 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": "stripe-disputes-to-respond/<date of this run, e.g. 2026-10-16>",
  "item_key": "stripe-disputes-to-respond",
  "project": "Acme",
  "source": "Stripe disputes to respond to skill",
  "kind": "action",
  "title": "4 disputes need evidence, the first due in 3 days",
  "summary": "**4** open disputes worth **$1,386** need a response. The $612 fraud claim is due Oct 20 and has no evidence saved yet.",
  "action": {
    "request": "Who gathers the evidence for these 4 disputes?",
    "options": [
      "Support takes all 4",
      "Fulfillment takes shipping",
      "I will do it",
      "Accept the $49 one"
    ]
  },
  "widget": [
    {
      "type": "Table",
      "props": {
        "columns": [
          "Customer",
          "Amount",
          "Reason",
          "Due",
          "Evidence"
        ],
        "align": [
          "left",
          "right",
          "left",
          "left",
          "left"
        ],
        "rows": [
          [
            "j.martin@example.com",
            "$612.00",
            "Fraudulent",
            "Oct 20",
            "None"
          ],
          [
            "orders@northwind.example",
            "$480.00",
            "Product not received",
            "Oct 23",
            "Tracking saved"
          ],
          [
            "li.chen@example.com",
            "$245.00",
            "Subscription canceled",
            "Oct 27",
            "None"
          ],
          [
            "k.okafor@example.com",
            "$49.00",
            "Duplicate (inquiry)",
            "Oct 29",
            "None"
          ]
        ]
      }
    },
    {
      "type": "Callout",
      "props": {
        "text": "The $49 duplicate is still an inquiry: answering it now can keep it from becoming a chargeback.",
        "tone": "info"
      }
    }
  ]
}
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: Daily at 9:00 (`0 9 * * *`). From cron, with the agent signed in on that machine:

```bash
0 9 * * * claude -p "Run the stripe-disputes-to-respond 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/stripe-disputes-to-respond#schedule

Source: https://briefkin.dev/skills/stripe-disputes-to-respond

What your agent needs

  • Read access to disputes and charges: the Stripe MCP server (https://mcp.stripe.com) with OAuth or a restricted key as the bearer token, which reads them through its list_disputes and search tools, or a restricted key with Disputes and Charges set to Read in STRIPE_API_KEY.

  • Who on the team can gather evidence (support, fulfillment, finance), so the card can offer them as answers.

Connecting Stripe

  1. Stripe MCP server, Stripe's official remote server at https://mcp.stripe.com. Add Stripe from Claude's connector directory, or run claude mcp add --transport http stripe https://mcp.stripe.com/, and sign in with OAuth; clients without OAuth can send a restricted key as a Bearer token.
  2. The Stripe CLI. Run stripe login once, then stripe subscriptions list --status active --limit 100 or stripe invoices list --status open. The CLI defaults to test mode, so add --live for real data.
  3. API with a restricted key. Create a restricted key (rk_live_...) in the Dashboard under Developers > API keys, store it in STRIPE_API_KEY, and give it Read on the resources the report touches: Subscriptions, Invoices, Charges, Customers, PaymentIntents, Balance and Disputes.

Never hand an agent the secret key (sk_live_...); a restricted key with Read where needed and None everywhere else can't refund, cancel or charge anything.

Questions

Can it submit the evidence?

Not in this skill. Updating a dispute with evidence sends it to the bank unless submit=false is set, and a dispute gets one submission, so the skill only reads. Gather the files, then submit from the Dashboard where you can review the whole response.

What is the difference between an inquiry and a dispute?

An inquiry (warning_needs_response) is the cardholder's bank asking for information before a formal chargeback, and no funds have been withdrawn yet. Left unanswered, it can turn into a dispute (needs_response), where the amount and a fee are already taken from your balance.

Do I need Briefkin to use the Stripe disputes to respond to 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 asks you, 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 Stripe pages themselves: the agent opens dashboard.stripe.com/disputes 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.

Answer your agents from one board, or from Telegram

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.