The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's installation process involves downloading scripts and configuration files from the author's GitHub repository (github.com/fourteenwm/ppc-ai-skills). These are vendor-owned resources from a well-known service and are documented as part of the setup.
[COMMAND_EXECUTION]: The workflow involves the local execution of Python scripts that interact with various cloud services. These scripts manage data gathering and sheet writing tasks essential to the skill's primary function.
[PROMPT_INJECTION]: The scrape_website_firecrawl.py script identifies and ingests untrusted text from external websites to extract business data via an LLM. Ingestion points: Scraped markdown content from the homepage, services, and about pages of the target business website. Boundary markers: The LLM prompt uses structured headers to separate instructions from data, although it does not use unique isolation delimiters. Capability inventory: The LLM output is restricted to a structured JSON format and is used solely for data population; it does not trigger command execution or filesystem modifications. Sanitization: The ingested website content is truncated to the first 10,000 characters to prevent potential context-overflow or denial-of-service attempts.

rsa-single-account