The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates web scraping by instructing the user to set up the Firecrawl MCP server using npx -y firecrawl-mcp. Firecrawl is a well-known service for automated web data extraction, and this is a standard installation method for its tools.
[REMOTE_CODE_EXECUTION]: The recommended configuration for the agent includes running the firecrawl-mcp package via npx, which executes remote code in the local environment. This is a functional requirement to enable the skill's advanced scraping capabilities.
[COMMAND_EXECUTION]: The skill uses the Bash tool to perform project setup and file generation, such as creating directories for research and writing the final website code (HTML, CSS, JS).
[PROMPT_INJECTION]: The skill ingests data from external websites (clients and competitors), which is a functional requirement for its purpose. This creates a surface for indirect prompt injection.
Ingestion points: Scraping client and competitor content via Firecrawl and WebFetch (SKILL.md, Phases 1 and 2).
Boundary markers: Scraped data is saved to Markdown files without specific delimiters or instructions to ignore embedded instructions.
Capability inventory: The agent utilizes Bash and Write tools to generate directories and code files based on the processed research.
Sanitization: No explicit sanitization or filtering of external content is described before the data is used for report and code generation.
[SAFE]: The skill follows security best practices for secret management by guiding users to store their API keys in a local environment configuration rather than hardcoding them. The behavior of the skill is consistent with its stated purpose of building research-backed websites.

website-intelligence