Claude Skills
Overview
Skill Use Cases
Reference

Shopping Assistant Skill


The Shopping Assistant skill (skill-shopping) turns Claude Code, Claude Desktop, OpenClaw, or any other LLM client into a personal shopping advisor. It runs a systematic 10-phase research workflow — from requirements gathering through review analysis to a final shortlist of 3-4 best options with a clear recommendation — using web search to gather real product data, ratings, reviews, and prices.

The point is to hit the sweet spot between under-researching (buying the first thing you see) and over-researching (days of analysis paralysis). The workflow is structured enough to converge on a decision and thorough enough to catch the blind spots casual browsing misses: newer disruptor brands, better-value models in the same lineup, and complaints that actually matter for your use case.

Get Started

1
Create an Account

Sign up for a free InfraNodus account to get access to the knowledge graph tools and your API key.

Create a Free Account
2
Get InfraNodus MCP
https://mcp.infranodus.com

Add this URL as a connector via Settings > Connectors in Claude Web / Desktop and authenticate with your InfraNodus API key via OAuth.

3
Install the Skill
/plugin marketplace add infranodus/skills
/plugin install skill-shopping@infranodus-skills

Run these two commands in the Claude Code prompt to install the skill as a plugin (use infranodus-all@infranodus-skills to install all our skills at once). Or use npx in your terminal: 

npx skills add infranodus/skills --skill skill-shopping -g -a claude-code -y

npx comes bundled with Node.js. For Claude Web / Desktop, ChatGPT, and OpenClaw, see the full installation instructions.

Didn't work? You can simply download skill-shopping.zip from our GitHub repository and add it to your LLM client manually — see the step-by-step instructions.

What the Skill Does

Once installed, prompts like these trigger the full research pipeline:

  • "What's the best 4K projector for a bright living room?" — runs the full workflow: requirements, candidates, comparisons, reviews, recommendation.
  • "Help me choose between these two laptops" — head-to-head comparison that surfaces differentiating features spec sheets don't show.
  • "I need a dishwasher under $600, delivered this week" — compresses the workflow for urgent purchases and prioritizes delivery and returns.
  • "Are there any newer brands I'm missing?" — a dedicated disruptor scan for entrants from the last 12-18 months that review roundups overlook.
  • "Just give me a recommendation" — quick mode: requirements, top 3 from the best review sites, recommendation.

Along the way it checks things most buyers skip: review volume vs rating (2,000 reviews at 4.2 beats 15 at 4.8), whether negative reviews are relevant to your use case, resale value, warranty, and return policies. It adapts depth to stakes — a $15 phone case gets a quick pass, a $1,500 purchase gets the full treatment.

How It Works: The 10 Research Phases

PhaseWhat happens
1. DiscoverWhat do you need, where will you buy it, budget, timeline, brand exclusions.
2. DefineBuild a prioritized feature list from your requirements, shopping-site filters, review pain points, and your deeper underlying needs. You rate each as must-have / important / nice-to-have.
3. SurveyFirst list of 8-12 candidates from cross-referenced review sites, plus a disruptor scan for new brands.
4. Compare"A vs B" comparisons reveal hidden features; competitors that reviewers benchmark against get added to the list.
5. OptimizeCheck each brand's lineup for a cheaper near-equivalent or a slightly pricier model with a key feature.
6. ExpandDeliberately break brand anchoring — alternatives from other brands, and even other product categories.
7. NarrowFinal shortlist of 3-4 options mapped against your priorities.
8. ReviewDeep dive into 5-star and 1-2-star reviews; irrelevant complaints filtered out, systematic issues flagged.
9. StretchOne last check 20-30% above and below the price range for genuine value shifts.
10. DecideResale value, delivery, warranty, support — then a clear recommendation with trade-offs stated.

InfraNodus Tools for Market Gap Analysis

The skill works standalone with web search, but when the InfraNodus MCP server is connected it starts each key phase with a graph analysis that surfaces patterns sequential web searches miss:

  • Define — analyze_related_search_queries reveals the features and brands real buyers search for, often replacing 5-10 individual web searches.
  • Survey — search_queries_vs_search_results finds what people want but can't find: the demand-supply gap where underrated products and disruptors live.
  • Compare — generate_content_gaps on combined review content finds blind spots that reading reviews linearly would miss.
  • Expand — generate_research_ideas with transcend mode checks whether a completely different solution serves your underlying need better (e.g. an OLED TV instead of a projector, if the real need is dark-room immersion).

Install the Skill

Claude Code

Download the skill into your global skills folder:

cd ~/.claude/skills
curl -L -o skill-shopping.zip https://github.com/infranodus/skills/releases/latest/download/skill-shopping.zip
unzip skill-shopping.zip -d skill-shopping

Then just describe what you want to buy — Claude will invoke the skill when relevant.

Claude Web / Desktop

Go to Settings > Capabilities, enable Code execution and file creation, scroll to the Skills section, and upload the skill-shopping.zip file.

OpenClaw

Run in the OpenClaw chat:

install this skill: https://github.com/infranodus/skills/releases/latest/download/skill-shopping.zip

See the main installation guide for step-by-step instructions per LLM client.

What's Next

Get the InfraNodus MCP