mova-complaints-handler

Submit a customer complaint for EU-compliant AI classification and human-in-the-loop handling decision via MOVA. Handles compensation claims, regulator threats, fraud signals, and repeat customer cases with mandatory human review and a full signed audit trail.

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "mova-complaints-handler" with this command: npx skills add mova-compact/mova-complaints-handler

Contract Skill — A ready-to-use MOVA HITL workflow. Requires the openclaw-mova plugin.

MOVA EU Consumer Complaints Handler

Turn any customer complaint into a structured, auditable handling decision — with AI triage, EU-compliant mandatory human review, and a signed audit receipt.

Why use this skill

  • AI classifies in seconds — sentiment flags, product risk, regulator threat detection, repeat customer check
  • Mandatory human gate — compensation claims, regulator threats, high-risk products, and fraud signals always route to a human officer
  • Full audit trail — every decision is timestamped, signed, and stored in an immutable MOVA audit journal (EBA/ESMA-ready)
  • Draft response hint — AI suggests tone and content for the customer response

Requirements

Plugin: MOVA OpenClaw plugin must be installed in your OpenClaw workspace.

Quick start

Tell your agent:

"Handle complaint CMP-2026-1002 from customer CUST-992 — product: investments, text: 'I want compensation for unauthorized charges and will escalate to the regulator.', prior complaints: CMP-0501, CMP-0604"

Demo

Complaint input AI classification & decision options Audit receipt & compact log

Test complaint

test_complaint_CMP-2026-1002.png

What the agent does

Step 1 — Submit complaint

Call tool mova_hitl_start_complaint with:

  • complaint_id (e.g. CMP-2026-1002), customer_id
  • complaint_text, channel (web/email/phone/chat/branch)
  • product_category (e.g. payments, investments, mortgage)
  • complaint_date (ISO date)
  • previous_complaints: optional array of prior complaint IDs
  • attachments: optional array of filenames
  • customer_segment: optional (retail/sme/corporate)
  • preferred_language: optional ISO 639-1 code

Step 2 — AI analysis output

Complaint:       CMP-2026-1002 — investments
Risk score:      0.85 / 1.0  (high)
Sentiment flags: compensation_claim, regulator_threat
Repeat customer: yes (2 prior complaints)
Findings:
  • COMP_CLAIM (high)           — Customer requests compensation for unauthorized charges
  • REGULATOR_THREAT (medium)   — Customer threatens regulator escalation
  • PRODUCT_RISK_INVEST (high)  — High-risk product category
Draft response hint: Acknowledge receipt, confirm investigation, outline compensation process
Recommended action: escalate ← RECOMMENDED

Step 3 — Human decision gate

OptionDescription
resolveSend standard response
escalateForward to complaints officer ← recommended
rejectMark as incomplete / invalid
regulator_flagFlag for regulator reporting

Call tool mova_hitl_decide with:

  • contract_id: from the response above — this is ctr-cmp-xxxxxxxx, NOT the complaint ID
  • option: chosen decision
  • reason: officer reasoning

Step 4 — Audit receipt

Call tool mova_hitl_audit with contract_id. Call tool mova_hitl_audit_compact with contract_id for the full signed event chain.

Audit log format (compact)

SeqEventDetails
1contract.startComplaint submitted
2step.ai_taskClassification completed
3step.verificationRisk snapshot passed
4step.decision_pointAwaiting human
5decision.humanOfficer selected option + reason
0metaContract finalized, status: completed

Data flows

  • Complaint text and metadata → MOVA Classification API (EU-hosted)
  • Human decision → MOVA Audit Journal (immutable, signed)
  • No data stored locally or sent to third parties

Connect your real CRM and policy systems

By default MOVA uses a sandbox mock. To route checks against your live CRM and policy engine, call mova_list_connectors with keyword: "crm".

Relevant connectors:

Connector IDWhat it covers
connector.crm.customer_lookup_v1Customer history and prior complaints from CRM
connector.policy.complaints_rules_v1Complaints handling rules by product/jurisdiction
connector.notification.email_v1Customer notification email

Call mova_register_connector with connector_id, endpoint, optional auth_header and auth_value.

Rules

  • Agent never makes HTTP requests directly
  • Agent never invents or simulates results — if a tool call fails, show the exact error
  • Use MOVA plugin tools directly — do NOT use exec or shell
  • CONTRACT_ID is ctr-cmp-xxxxxxxx from the mova_hitl_start_complaint response — NOT the complaint ID

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

Ephemeral Media Hosting

自動削除機能付き一時メディアホスティングシステム

Registry SourceRecently Updated
General

Ethereum Read Only

Foundry castを使用したウォレット不要のオンチェーン状態読み取り

Registry SourceRecently Updated
General

OpenClaw Memory

Manage, optimize, and troubleshoot the OpenClaw memory system — MEMORY.md curation, daily logs (memory/YYYY-MM-DD.md), memory_search tuning, compaction survi...

Registry SourceRecently Updated
General

ImageRouter

Generate AI images with any model using ImageRouter API (requires API key).

Registry SourceRecently Updated
2.6K2dawe35