The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a local Python script to perform API interactions. This is limited to the defined functionality and does not allow for arbitrary command injection.- [EXTERNAL_DOWNLOADS]: Network requests are directed to api.browseract.com, which is the official API domain for the tool's developer, browser-act. These communications are necessary for the skill's function.- [CREDENTIALS_UNSAFE]: The skill utilizes the BROWSERACT_API_KEY environment variable for authentication, which is a secure practice. It explicitly guides users to set this variable if it is missing.- [DATA_EXFILTRATION]: No sensitive local files or data are transmitted. The script only sends the target Amazon ASIN and marketplace URL to the scraping service.- [PROMPT_INJECTION]: The skill ingests untrusted data from external Amazon listings.
Ingestion points: Product data (titles, bullets, reviews) entering through scripts/amazon_listing_competitor_analysis.py.
Boundary markers: The SKILL.md defines a structured framework for Phase 2 analysis, though it lacks explicit technical delimiters for external data.
Capability inventory: The skill's capabilities are limited to the specific API script execution.
Sanitization: The skill does not perform specific sanitization, but relies on its structured analysis framework to focus on data facts rather than instructions.

amazon-listing-competitor-analysis-skill