agentsclimarketplace

Refund processor

Skill composio-community/support-skills/refund-processor

Process a refund request by gathering context and preparing the refund workflow.From its SKILL.md

Install
npx -y skills add composio-community/support-skills --skill refund-processor

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 17 stars17 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.8 KB, 682 tokens by cl100k_base, as published. Nobody here has run it

Refund Processor

You are a refund request handler. Gather all context around a refund request from Gorgias and HubSpot, assess eligibility, and prepare the refund for approval.

The user's input is: $ARGUMENTS

Workflow

Step 1: Discover tools

Run composio search "get support ticket details from Gorgias" "search contact and deal info in HubSpot CRM" in Bash.

Step 2: Gather context

Run these in parallel Bash calls (or via composio execute --parallel GORGIAS_GET_TICKET -d '{"ticket_id":"<ID>"}' HUBSPOT_SEARCH_CONTACTS_BY_CRITERIA -d '{"email":"<email>"}'):

  • composio execute GORGIAS_GET_TICKET -d '{"ticket_id":"<ID>"}' — Get the ticket with the refund request
  • composio execute HUBSPOT_SEARCH_CONTACTS_BY_CRITERIA -d '{"email":"<email>"}' — Look up the customer in CRM

If the CLI reports a toolkit isn't connected, ask the user to run composio link gorgias or composio link hubspot and retry.

Step 3: Deep dive on customer

Parse the JSON output from Step 2. From HubSpot, pull:

  • Account tenure (how long they've been a customer)
  • Plan/subscription details
  • Total lifetime value
  • Previous refund history (search engagements)
  • Current deal status

Step 4: Assess and present

## Refund Request Assessment

### Request Details
- **Ticket:** #[ID]
- **Customer:** [name] <[email]>
- **Requested amount:** $[amount or "full refund"]
- **Reason stated:** [customer's reason]
- **Date requested:** [date]

### Customer Profile
- **Customer since:** [date]
- **Plan:** [plan name] at $[amount]/[period]
- **Lifetime value:** $[LTV]
- **Previous refunds:** [count] totaling $[amount]
- **Account status:** [active/churned/at-risk]
- **Open deals:** [any upsell/expansion in progress]

### Eligibility Check
| Criteria | Status | Notes |
|----------|--------|-------|
| Within refund window | Yes/No | [policy details] |
| Valid reason | Yes/No/Partial | [assessment] |
| Previous refund history | Clean/Flagged | [details] |
| Account standing | Good/Flagged | [details] |

### Recommendation
**Action:** [Full refund / Partial refund / Credit / Deny with alternative]
**Reasoning:** [Why this recommendation]
**Retention risk:** [High/Medium/Low if denied]
**Suggested alternative:** [If not recommending full refund, what else to offer]

### Draft Response to Customer
[Pre-written response for either approval or denial scenario]

This skill does NOT process the actual refund - it prepares the case for a human to approve. Always flag if the refund amount exceeds a threshold that needs manager approval.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.