Global Price Comparison

Discover and compare the same product across multiple countries and source types (official stores, marketplaces, retailers) using Brave and/or Tavily web sea...

Security Scan

The skill's code and instructions are coherent with its stated purpose (discover product links via Brave/Tavily and normalize prices to USD); it requires only optional search API keys and makes expected network calls.

Purpose & Capability

Name/description (global product price discovery and USD normalization) match the included script and docs: the script performs discovery via Brave/Tavily and converts local prices to USD using a public FX API. No unrelated capabilities or credentials are requested.

Instruction Scope

SKILL.md instructs running the provided script and filling a CSV; the discover command requires BRAVE_API_KEY and/or TAVILY_API_KEY (documented). The runtime instructions do cause outbound HTTP requests to Brave, Tavily, and the public FX API (open.er-api.com) which is consistent with discovery/normalization. This is expected, but you should be aware the skill will send product queries and URLs to those external services when discovery is used.

Install Mechanism

No install spec (instruction-only with a bundled Python script). Nothing is downloaded or written by an installer step. The script runs locally and is the only code artifact.

Credentials

The skill does not declare required env vars in registry metadata, but the documentation and script reference optional BRAVE_API_KEY and TAVILY_API_KEY for discovery. Requesting these service API keys is proportionate to the discovery feature. There are no unrelated or excessive credential requests.

Persistence & Privilege

always is false and the skill does not request persistent/system-wide privileges. It does not attempt to write or modify other skill configurations based on the provided files.

Assessment

This skill appears to do what it says: search public search APIs for candidate product links and convert prices to USD. Before installing/providing API keys, consider: (1) Only supply BRAVE_API_KEY and/or TAVILY_API_KEY if you trust those services and are comfortable that the agent will make search requests (keys are sent to those APIs). Use limited-scope or revocable keys where possible. (2) The script makes outbound requests to open.er-api.com for exchange rates; verify you are comfortable with that external dependency. (3) Review the bundled Python script locally if you want to verify it does not transmit data beyond search/discovery results. (4) Because the agent can invoke skills autonomously by default, if you do provide API keys expect the agent to perform network queries without prompting during discovery flows—restrict or rotate keys if needed. If you want higher assurance, request a full file diff or run the script in an isolated environment first.

Use this skill to run a global same-product price scan and output a USD-normalized comparison.

Quick start

# 1) Create a starter CSV template
python scripts/global_price_compare.py template --out /tmp/offers.csv

# 2) (Optional) discover candidate links by country/source type
# Uses Brave and Tavily when available (BRAVE_API_KEY / TAVILY_API_KEY)
python scripts/global_price_compare.py discover \
  --product "iPhone 16 Pro 256GB" \
  --countries US,JP,DE,UK \
  --source-types official_store,marketplace,electronics_retailer \
  --engine all \
  --out /tmp/discover.json

# 3) Fill /tmp/offers.csv with verified offers, then compare
python scripts/global_price_compare.py compare \
  --input /tmp/offers.csv \
  --format markdown

Workflow

1. Define exact product variant (model/storage/spec).
2. Run discover to get candidate URLs by market and source type.
3. Add verified offers to CSV (product,country,currency,source_type,source_name,price,url).
4. Run compare to normalize all prices to USD and rank best/worst.
5. Share markdown/JSON/CSV output.

Commands

Template

python scripts/global_price_compare.py template --out /tmp/offers.csv

Discover candidate links

python scripts/global_price_compare.py discover \
  --product "PlayStation 5 Slim" \
  --countries US,JP,DE \
  --source-types official_store,marketplace,electronics_retailer \
  --engine all \
  --count 3 \
  --format markdown

Notes:

• discover supports --engine brave|tavily|all (default all).
• For all, it uses Brave first and only falls back to Tavily when Brave fails or returns no results.
• Set one or both keys as needed: BRAVE_API_KEY, TAVILY_API_KEY.
• Source types are intentionally generic (not local-store specific).

Compare offers in USD

python scripts/global_price_compare.py compare \
  --input /tmp/offers.csv \
  --format markdown \
  --out /tmp/comparison.csv

Outputs include:

• ranked offers by USD price
• delta vs best offer
• spread in USD and %

Defaults

Default country set:

• US, UK, JP, DE, FR, CA, AU, SG, HK, TW

Default source types:

• official_store
• marketplace
• electronics_retailer
• general_retailer

Data/quality guardrails

Before finalizing recommendations:

• confirm same model/variant
• confirm tax/shipping basis consistency
• separate new vs refurbished/used
• keep URL + timestamp evidence

Detailed schema + guardrails:

• references/data-shape-and-source-types.md

Resources

• Script: scripts/global_price_compare.py
• Reference: references/data-shape-and-source-types.md