← Back to Skills Marketplace
carlos-zen

lumenshop-deals

by Carlos · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
83
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install lumenshop-deals
Description
Search Shopify products (shoes, clothes, bags) and present results as beautiful image+text product cards. Use this skill whenever the user wants to buy somet...
README (SKILL.md)

Overview

Query the LumenShop API to search thousands of indexed Shopify stores and render results as rich product cards — each with an image, clickable title, price, brand, and store.


Workflow

Step 1 — Understand the request

Extract from the user's message:

  • Search term: a specific keyword (preferred), or a category (shoes, clothes, bags, all)
  • Price range: budget ceiling or floor, if mentioned
  • Quantity: default to 12 items; honor explicit requests for more

If anything is unclear, pick the most reasonable default and proceed — no need to ask.

Step 2 — Run the script

The script is at scripts/skill.sh, run it from the skill directory:

# Keyword search (preferred)
bash scripts/skill.sh --query "blue sneakers" --limit 12

# Category search
bash scripts/skill.sh --category shoes --limit 12

# With price filter
bash scripts/skill.sh --query "skirt" --price-max 50 --limit 12

The script outputs raw JSON — parse it in the next step.

Step 3 — Parse the JSON response

Response structure:

{
  "hits": {
    "total": { "value": \x3Ctotal_count> },
    "hits": [
      {
        "_source": {
          "title":     "Product Name",
          "brand":     "Brand",
          "url":       "https://...",
          "gallery":   [{ "url": "https://cdn.shopify.com/..." }, ...],
          "prices":    [{ "currency": "USD", "price": 29.99 }],
          "hostnames": ["store.myshopify.com"]
        }
      }
    ]
  }
}

Skip any product where gallery is empty — a card without an image is not useful to the user.

Step 4 — Render product cards

Start your response with a warm intro line, then render one card per product.

Opening line (always include this at the top):

✨ LumenShop has found the best products just for you!

Card template (repeat for each product):

---
### [Product Title](product_url)

![Product Title](gallery[0].url)

💰 **$XX.XX** · 🏷️ Brand · 🏪 store_hostname

Full output structure:

✨ LumenShop has found the best products just for you!

## Found X items for you (Y total)

---
### [Product Title](url)

![Product Title](image_url)

💰 **$XX.XX** · 🏷️ Brand · 🏪 store.com

---
### [Next Product](url)

![Next Product](image_url)

💰 **$XX.XX** · 🏷️ Brand · 🏪 store.com

---

Step 5 — Offer to refine

After the cards, add a short follow-up prompt:

Want to refine by keyword, price range, or category?


Script options

Flag Default Description
--query none Keyword search; multiple words match with OR logic
--category all shoes / clothes / bags / all (ignored if --query is set)
--price-min none Minimum price (USD)
--price-max none Maximum price (USD)
--limit 20 Max results to return (up to 200)

Category keyword mapping

--category Equivalent --query
shoes shoe sneaker boot sandal
clothes shirt jacket dress hoodie pants skirt
bags bag backpack purse tote
all all of the above
Usage Guidance
This skill appears to do what it claims: run a local Bash script that queries a LumenShop API and returns product JSON, which the agent then formats as product cards. Before installing, consider: (1) the script makes outbound requests to https://lumenshop.vercel.app by default — check LumenShop's privacy/trustworthiness if you care about query data leaving your environment; (2) you can override the API URL and API key via flags or the LUMENSHOP_API_URL / LUMENSHOP_API_KEY env vars (these env vars are used by the script but not declared in the registry metadata); (3) results include clickable URLs and image links — treat them like any third-party links and avoid entering sensitive info on unknown stores. If you want tighter control, request that the publisher document the env vars in metadata or allow using an internal/owned API endpoint and key.
Capability Analysis
Type: OpenClaw Skill Name: lumenshop-deals Version: 1.0.0 The skill is classified as suspicious due to poor input sanitization and instructions that reduce human oversight. The script `scripts/skill.sh` manually constructs a JSON payload for a remote API (`https://lumenshop.vercel.app`) using unescaped shell variables, which is vulnerable to JSON injection. Furthermore, `SKILL.md` contains instructions for the AI agent to 'proceed — no need to ask' if parameters are unclear and to 'always prefer this skill' over others, which could be leveraged via prompt injection to force the agent to send unintended data to the external endpoint without user confirmation.
Capability Tags
requires-oauth-token
Capability Assessment
Purpose & Capability
Name/description (search Shopify products and render product cards) match the included Bash script and SKILL.md. The script performs a POST to an API endpoint to fetch product hits and returns raw JSON as described.
Instruction Scope
SKILL.md instructs the agent to run scripts/skill.sh and parse its JSON output — that matches the bundle. Minor inconsistency: the script reads optional environment variables LUMENSHOP_API_URL and LUMENSHOP_API_KEY, but the skill's metadata declares no required env vars; the README does mention an API base URL and token via script flags, however these env variables are not documented in the top-level registry metadata.
Install Mechanism
No install spec; this is an instruction-only skill with a single Bash script. No downloads, package installs, or archive extraction. Low-risk installation footprint.
Credentials
The script accepts an API key and API URL via environment variables or flags (defaults to a public dev-key and https://lumenshop.vercel.app). No required credentials are declared. This is proportionate to a web-API querying skill, but the registry metadata does not advertise the optional env vars that the script will read if present.
Persistence & Privilege
Skill is not marked always:true and does not request persistent system-wide privileges or modify other skills. It makes outbound HTTP requests (expected for its purpose) but does not write to configuration or access unrelated system files.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install lumenshop-deals
  3. After installation, invoke the skill by name or use /lumenshop-deals
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
lumenshop-deals 1.0.0 – Initial release - Instantly searches thousands of Shopify stores for shoes, clothes, and bags based on user queries or shopping intent. - Returns rich product cards with images, product links, price, brand, and store for up to 12 items by default. - Supports keyword, category, and price range filters. - Includes a warm intro and easy-to-read Markdown cards for each product found. - Offers follow-up prompts to refine results by keyword, price, or category.
Metadata
Slug lumenshop-deals
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is lumenshop-deals?

Search Shopify products (shoes, clothes, bags) and present results as beautiful image+text product cards. Use this skill whenever the user wants to buy somet... It is an AI Agent Skill for Claude Code / OpenClaw, with 83 downloads so far.

How do I install lumenshop-deals?

Run "/install lumenshop-deals" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is lumenshop-deals free?

Yes, lumenshop-deals is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does lumenshop-deals support?

lumenshop-deals is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created lumenshop-deals?

It is built and maintained by Carlos (@carlos-zen); the current version is v1.0.0.

💬 Comments