pangolinfo-amazon-daily-competitor-radar

Name/description (Amazon competitor radar) align with the bundled scripts: ai_serp.py, amazon_scraper.py, amazon_niche.py, and wipo.py. The included WIPO trademark lookup is coherent with competitor/brand monitoring. Requested env vars are the Pangolinfo credentials needed to call the APIs hosted at scrapeapi.pangolinfo.com, matching the stated purpose.

SKILL.md instructs the agent to run the bundled Python clients and to use the Pangolinfo endpoints. The scripts only reference the declared environment variables and an optional cache file (~/.pangolinfo_api_key) and do not attempt to read other system paths or unrelated credentials. Network calls are directed to pangolinfo domains described in the docs.

No install spec or external downloads are present; the skill is delivered as bundled Python scripts (zero-dependency) which is low risk. There is no code that pulls arbitrary binaries or executes downloads from shorteners/personal servers.

The skill requires PANGOLINFO_API_KEY, PANGOLINFO_EMAIL, and PANGOLINFO_PASSWORD but the docs and scripts accept either an API key OR email+password (the metadata note reflects this). This is reasonable, but users should be aware that providing email+password will allow the scripts to exchange those credentials for a long-lived API token (and optionally cache it if the user enables caching).

always is false and the skill does not request elevated system privileges. The only persistent storage behavior is an optional, user-controlled cache file (~/.pangolinfo_api_key) when the user explicitly enables caching. Autonomous invocation (model-invocation enabled) is the platform default and not itself unusual here.