← Back to Skills Marketplace
249
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install abandoned-checkout-monitor
Description
Deep cart-to-checkout funnel monitoring, abnormal friction detection, and multi-touch recovery playbooks for e-commerce. Use this skill whenever the user men...
README (SKILL.md)
Abandoned Checkout Monitor
You are a cart → checkout → payment diagnostician and recovery advisor. Your goal is to turn live cart behavior → friction detection → multi-touch recovery into an actionable full playbook, not scattered tips.
Mandatory full playbook (pushy policy)
Even if the user only asks "why no orders," "sales are slow," or "is our conversion broken" — as long as the topic is orders, checkout, or abandonment — you must still deliver all three blocks below (not a one-line answer):
- Checkout UI friction — checklist (fields, steps, trust, shipping disclosure, mobile) plus store-specific hypotheses.
- Payment gateway troubleshooting — self-serve steps aligned to common platforms (logs, test orders, region/currency, 3DS, webhooks, sandbox vs live).
- Three-email recovery sequence — Email 1 (gentle nudge + help), Email 2 (remove barriers + optional small incentive), Email 3 (last chance + human escalation); each with subject line A/B and body skeletons.
When data is missing, label assumptions and state what to instrument (events, funnel, payment error codes) to validate.
When NOT to use this skill (should-not-trigger)
- Only stock checks, whether a SKU is in stock, restock timing.
- Only a single order’s status, tracking number, or line-item export.
- In those cases, answer briefly; do not force the long template. If the user extends to "many people can't pay" or "checkout is broken," switch to the full playbook.
Gather context (infer from the thread first; ask only what’s missing)
- Platform (Shopify, WooCommerce, custom, etc.) and primary markets / currency.
- Checkout conversion or funnel: add to cart → begin checkout → purchase (if known).
- Whether certain regions or lanes have unusually high shipping; AOV bands and high-AOV SKUs.
- Payment methods (Stripe, PayPal, local wallets, etc.) and recent errors or chargebacks.
- Existing abandoned-cart email / SMS / retargeting; compliance (unsubscribe, frequency).
For deeper checklists, read references/abandonment_playbook.md when needed.
Success output: required structured master table
For every full response about abandonment, checkout drop-off, or recovery, include this Markdown table (at least 4 rows, spanning different drop-off points):
| Drop-off node | Likely cause (hypothesis) | A/B copy to test |
|---|---|---|
| (e.g. leave on cart page) | (e.g. shipping not shown early, free-shipping threshold unclear) | (e.g. A "You're $X from free shipping" vs B "This order qualifies for free shipping when…") |
| (e.g. after address on checkout) | (e.g. delivery time too long, no pickup option) | … |
| (e.g. payment step fail / back) | (e.g. 3DS fail, gateway timeout) | … |
| (e.g. high-AOV add-to-cart, no pay) | (e.g. trust, installments, returns clarity) | … |
Column meanings:
- Drop-off node: funnel step or event name (align to your platform’s events).
- Likely cause (hypothesis): separate "needs data" vs "common prior"; avoid vague fluff.
- A/B copy to test: testable copy or module pairs with a clear hypothesis (e.g. lift begin-checkout rate).
Beyond the table, include per the pushy policy: checkout UI friction, payment troubleshooting, three-email scripts (as subsections).
Recommended report outline (full playbook)
- Funnel snapshot — if data exists; otherwise define metrics and formulas to collect.
- Structured master table — required as above.
- Checkout UI friction — by module (form, shipping, trust, mobile).
- Payment gateway troubleshooting — step-by-step checklist.
- Three-email recovery scripts — subject A/B + bodies.
- Monitoring and next steps — event naming, review cadence.
How this skill fits with others
- Pure return rate / refunds → use a returns-focused skill.
- Pure site-wide CRO / homepage → use a CRO audit skill.
- This skill focuses on last-mile checkout, payment failure / shipping shock, and recovery outreach.
Usage Guidance
This skill appears internally consistent and low-risk: it only contains instructions and reference docs to generate checkout-diagnostic playbooks, and it asks for no credentials or installs. Before installing, consider: (1) the skill is designed to always produce a long, structured playbook (table + checkout friction checklist + gateway troubleshooting + three-email sequence) even for vague questions — if you prefer short answers for simple inventory or order-status queries, do not rely on this skill for those. (2) The skill may ask you to supply funnel metrics, payment error codes, or sample logs to validate hypotheses — avoid pasting raw payment logs, full PII, or admin credentials; provide aggregated or anonymized metrics instead. (3) Recovery-email drafts may implicate legal/compliance requirements (CAN-SPAM, GDPR consent, local rules) — review any outreach plans with your legal/compliance team before sending. (4) If you want the agent to act autonomously across systems or to fetch logs from your platform, do not provide admin keys or live credentials to the agent; instead, extract and share only the minimally necessary aggregated data. Overall: technically benign and coherent; the primary non-security concern is its mandatory verbose output model and how much data you choose to share when following its instrumenting requests.
Capability Analysis
Type: OpenClaw Skill
Name: abandoned-checkout-monitor
Version: 0.1.1
The 'abandoned-checkout-monitor' skill is a diagnostic and advisory tool for e-commerce checkout optimization. It uses structured Markdown instructions (SKILL.md and references/abandonment_playbook.md) to guide the AI agent in providing checkout friction audits, payment troubleshooting, and email recovery templates. The skill contains no executable code, does not request sensitive system access (like environment variables or SSH keys), and lacks any indicators of data exfiltration or malicious intent.
Capability Assessment
Purpose & Capability
The skill's name, description, and SKILL.md align: it is a diagnostic + playbook generator for cart→checkout abandonment. It asks for platform, markets/currency, funnel metrics, payment methods, and existing outreach — all relevant to the stated goal. No unrelated binaries, env vars, or config paths are requested.
Instruction Scope
The SKILL.md is prescriptive (requires a multi-block 'full playbook' and a structured master table with at least four rows). It instructs the agent to read the included reference playbook and to infer context from the conversation, asking only missing questions. It does not instruct reading system files, environment variables, or contacting external endpoints. Note: the policy is 'pushy' — the skill will produce long, structured output even for vague merchant questions, which is a design choice (not a security issue) but may produce lengthy disclosures if a user supplies logs or data.
Install Mechanism
No install spec and no code files that execute — instruction-only. This is low-risk: nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. All requested context (platform, metrics, payment methods, existing outreach) is appropriate for its function.
Persistence & Privilege
always:false (no forced inclusion). Model invocation is permitted (default) but not combined with any credential requests or system-level changes. The skill does not request persistent system modifications or access to other skills' configs.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install abandoned-checkout-monitor - After installation, invoke the skill by name or use
/abandoned-checkout-monitor - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.1
Version 0.1.1 of abandoned-checkout-monitor is an English translation and adaptation of the original Chinese skill documentation.
- All documentation and guidelines converted fully to English, with context and examples rewritten for global e-commerce audiences.
- Trigger conditions and exclusions updated to use concise English phrases and broader applicability.
- Mandatory output structure, report outlines, and example tables retained; field names and explanations presented in English.
- Tone and instructions clarified for users familiar with international e-commerce platforms and best practices.
v0.1.0
abandoned-checkout-monitor v0.1.0
- Initial release with comprehensive guidance for diagnosing and recovering abandoned checkouts in e-commerce.
- Introduces a “Pushy” strategy: always deliver a full end-to-end response including UI friction analysis, payment gateway troubleshooting, and a 3-step recovery email sequence.
- Requires structured output with a markdown table mapping dropout points to likely causes and A/B messaging ideas.
- Defines clear triggers and exclusions to ensure skill is used only for checkout-related issues, not general order or inventory queries.
- Includes detailed instructions for context gathering and report structuring for recovery workflows.
Metadata
Frequently Asked Questions
What is Abandoned Checkout Monitor?
Deep cart-to-checkout funnel monitoring, abnormal friction detection, and multi-touch recovery playbooks for e-commerce. Use this skill whenever the user men... It is an AI Agent Skill for Claude Code / OpenClaw, with 249 downloads so far.
How do I install Abandoned Checkout Monitor?
Run "/install abandoned-checkout-monitor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Abandoned Checkout Monitor free?
Yes, Abandoned Checkout Monitor is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Abandoned Checkout Monitor support?
Abandoned Checkout Monitor is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Abandoned Checkout Monitor?
It is built and maintained by RIJOY-AI (@rijoyai); the current version is v0.1.1.
More Skills