Reads one support ticket, order, and payment, evaluates the request, issues only the approved bounded refund, updates the customer ticket, and leaves financial and support proof.
Designed for
Support and billing teams resolving refund requests across Zendesk, Shopify, and Stripe
Operating goal
Resolve one evidence-backed refund request without duplicate financial or customer-facing actions.
4 steps from trigger to verified handoff, with success and failure paths.
1 MCP layer and 5 connected tools with explicit auth and risk levels.
3 guardrails, 3 evals, and 2 harnesses before production use.
Builds one cross-system refund evidence pack.
Grounded reasoning model
Applies policy and prepares exact bounded resolution fields.
Conservative policy model
Issues one approved Stripe refund and preserves its receipt.
Structured output model
Updates the ticket only after refund proof exists.
Customer-safe writing model
Loads the workflow goal, allowed actions, escalation policy, and output contract before the agent plans work.
A workflow skill that captures the operating contract, tool boundaries, and escalation rules for Refund Resolution Team.
Centralizes high-risk action checks for writes, secrets, customer data, billing, deploys, and public communications.
Exposes task resources, prompt templates, connector tools, and audit records behind a permission-aware boundary.
Read the ticket, order, and payment and reconcile identity and state.
Choose the bounded refund fields and draft the exact customer reply.
Issue the exact approved refund once and preserve the support fields.
Add the approved reply and ticket status after Stripe proof exists.
Open it in Codelit, refine it with the agent chat, then generate the architecture or product plan from the same workflow spec.
Open in Agent WorkflowReads one Jira project, converts a request into a scoped work item, pauses on the exact fields, creates it once, and returns the Jira issue proof.
Reads one Linear team's queue, turns incident evidence into a concise follow-up, pauses on the exact issue, creates it once, and returns Linear proof.
Find a customer ticket, prepare a grounded resolution, and add the approved comment or status update. The team reads one bounded scope, shows the exact change, executes it once after approval, and returns provider evidence.