All starter kits

Dispute Resolution

From user complaint to sealed proof card

Your support queue receives a complaint. Currently a human reads it, emails back for screenshots, and manually decides whether to refund. Wrap the intake with Patchr: it builds an evidence map, identifies contradictions, packages a proof card, and returns a recommended action — all before the ticket reaches a human.

Run in sandbox Workflow template

Agent route

ResolveProxy→ proof card
Before Patchr

Manual process, no audit trail, inconsistent outcomes.

// Before: manual, no audit trail
async function handleComplaint(ticket) {
  const messages = await fetchTicketMessages(ticket.id);
  // ❌ No evidence map — agent reads and decides manually
  // ❌ No contradiction check
  // ❌ No tamper-evident proof card
  if (messages.some(m => m.text.includes("mismatch"))) {
    await escalateToHuman(ticket.id);
  }
}
After Patchr — Node.js

One SDK call. Evidence map, proof card, and recommended action returned.

import { PatchrClient } from "@patchr-core/sdk";

const patchr = PatchrClient.fromEnv();

async function handleComplaint(ticket) {
  const run = await patchr.runOrchestrator({
    clientId:       "support-ops",
    channel:        "api",
    conversationId: `conv_${ticket.id}`,
    request:        ticket.summary,
    attachments:    ticket.screenshots.map(url => ({
      name: "evidence.jpg",
      mime: "image/jpeg",
      url,
    })),
  });

  // ✅ Evidence map built, contradictions flagged
  // ✅ Proof card sealed with confidence + open questions
  // ✅ Recommended restraint: "escalateBeforeActing" | "proceedSafely"
  console.log(run.status);           // "ready" | "NEEDS_INPUT"
  console.log(run.proofCard?.recommendedRestraint);
  console.log(run.proofCard?.cite);  // portable cite ref
}
After Patchr — Python

Same integration using the raw HTTP API with the httpx client.

import os, httpx

PATCHR_API = os.environ["PATCHR_API_BASE_URL"]
PATCHR_TOKEN = os.environ["PATCHR_API_TOKEN"]

def handle_complaint(ticket):
    resp = httpx.post(
        f"{PATCHR_API}/v1/orchestrator/run",
        headers={"Authorization": f"Bearer {PATCHR_TOKEN}"},
        json={
            "clientId":       "support-ops",
            "channel":        "api",
            "conversationId": f"conv_{ticket['id']}",
            "request":        ticket["summary"],
            "attachments": [
                {"name": "evidence.jpg", "mime": "image/jpeg", "url": url}
                for url in ticket.get("screenshots", [])
            ],
        },
    )
    run = resp.json()

    # ✅ Evidence map built, contradictions flagged
    # ✅ Proof card sealed — cite ref portable across systems
    print(run["status"])
    print(run.get("proofCard", {}).get("recommendedRestraint"))
    print(run.get("proofCard", {}).get("cite"))

Output shape

A consistent proof card — every run.

Whether you run a dispute audit or an agent governance check, the output shape is identical. External systems — dashboards, decision ledgers, compliance tools — consume the same format regardless of which workflow produced it.

evidenceStatus — sourceBound | verified | partial

confidence — 0–1 score from the evidence map

recommendedRestraint — proceedSafely | proceedWithCaution | escalateBeforeActing

cite — portable proof ref for cross-system ledgers

Proof card format

Sample output

{
  "status": "ready",
  "route": ["resolve"],
  "proofCard": {
    "title":                 "Audit this disputed claim",
    "evidenceStatus":        "sourceBound",
    "confidence":            0.74,
    "openQuestions":         ["contradictionsRequireReview"],
    "recommendedRestraint":  "escalateBeforeActing",
    "cite":                  "patchr:proofCard:card_...#v1"
  }
}

Ready to run it?

Audit disputed claim ZD-44291: the seller name and payment instructions on the invoice do not match. Evidence includes a transaction screenshot and the policy document.

Open in sandbox