Shopping

Help users make better purchase decisions — research, comparison, timing, and avoiding buyer's remorse.

Security Scan

An instruction-only shopping advisor whose requested capabilities and runtime instructions are consistent with its stated purpose and do not request credentials, installs, or system access.

Purpose & Capability

The name and description (shopping advice, comparison, timing) match the SKILL.md content. The skill's recommendations (review sources, price trackers, return policy checks) are coherent with a shopping advisor. It does not request unrelated binaries, credentials, or config paths.

Instruction Scope

SKILL.md is a set of heuristics and step-by-step guidance for asking user questions and researching external reviews/price trackers. It does not instruct the agent to read local files, environment variables, or transmit user secrets. It references external web sources (CamelCamelCamel, Honey, Reddit, YouTube) which is appropriate for research-oriented advice.

Install Mechanism

No install spec and no code files: the skill is instruction-only. That results in low install risk because nothing is written to disk or executed automatically by the platform.

Credentials

The skill declares no required environment variables, credentials, or config paths. There are no disproportionate requests for secrets or unrelated service tokens.

Persistence & Privilege

Flags are default (always:false, agent can invoke autonomously). The skill does not request persistent presence or modify other skills or system settings.

Assessment

This is a lightweight, instruction-only shopping advisor and appears internally consistent. Points to consider before installing: (1) Source/homepage is unknown — the content is generic and harmless, but lack of provenance means you may want to prefer skills from known authors. (2) The skill directs checking third-party websites (price trackers, forums); those sites may track you or have affiliate links — avoid sharing payment or account credentials. (3) The skill does not request any credentials or perform purchases itself, but if you ask the agent to take actions (buy, log in), treat those requests cautiously. If you want extra safety, confirm the agent will only provide recommendations and not submit forms or store sensitive information.

Before Recommending Products

• Ask budget range first — recommendations without budget waste time
• Ask use case specifically — "laptop" means different things for gaming vs email
• Ask what they've tried or owned — past experience reveals preferences
• Ask timeline — urgent need vs can wait for sale changes strategy
• One recommendation with reasoning beats list of options

Research Approach

• Check reviews from multiple sources — single source can be biased/paid
• Filter 1-star reviews for real issues — ignore "shipping was slow" complaints
• Look for long-term reviews — 6-month updates reveal durability
• Reddit/forums often more honest than YouTube — less sponsorship pressure
• Check if newer model coming — buying at end of cycle means faster obsolescence

When to Suggest Not Buying

• "I might need this someday" — future need isn't current need
• Upgrading something that works fine — marginal improvement, full price
• Buying to solve problem that isn't the product — new running shoes won't create running habit
• Emotional purchase after bad day — suggest waiting 48 hours
• Sale pressure: "70% off ends tonight" — if didn't need it yesterday, don't need it today

Price and Timing

• Track price history: CamelCamelCamel for Amazon, Honey for others — "sale" might be normal price
• Major sales: Black Friday, Prime Day, end of season — worth waiting if not urgent
• Refurbished/open-box: often 20-40% off, same warranty — underrated option
• Credit card price protection — check if card offers it before buying
• Price match policies — many stores match competitors, just ask

Comparison Framework

Factor	Questions
Must-haves	What features are non-negotiable?
Nice-to-haves	What would be bonus but not essential?
Deal-breakers	What would make you return it?
Total cost	Accessories, subscription, maintenance?
Longevity	How long until you need to replace?

Two good options? Pick the one easier to return.

Red Flags

• Too many 5-star reviews with similar language — likely fake
• Brand new product with hundreds of reviews — suspicious timing
• "Amazon's Choice" means nothing about quality — just algorithm pick
• Influencer discount codes — they profit from your purchase
• Pressure tactics: countdown timers, "only 2 left" — manufactured urgency

Return Policy Awareness

• Check return policy before buying — some categories no returns
• Keep packaging until sure — needed for returns
• Credit card extended return windows — some cards add 90 days
• Restocking fees on electronics — factor into decision
• "Final sale" means final — no exceptions

Category-Specific Guidance

• Electronics: Refresh cycles matter, buy early in cycle not end
• Clothing: Size charts vary wildly, read size reviews specifically
• Furniture: Measure twice, assembly difficulty in reviews
• Appliances: Check repair frequency ratings (not just features)
• Subscriptions: Calculate yearly cost, check cancellation ease

Post-Purchase

• Don't keep researching after buying — decision fatigue, regret spiral
• Found cheaper after purchase? Many stores price-match within window
• Problem with product? Contact support before leaving bad review
• Actually use the thing — purchase isn't the goal, use is