The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to download and execute the Ollama installation script directly from the official ollama.ai domain using curl -fsSL https://ollama.ai/install.sh | sh in guides/05-local-llms.md.
[COMMAND_EXECUTION]: Multiple guides, specifically guides/07-mcp-and-ide-plugins.md, instruct the user or agent to execute npx and uvx commands to run Model Context Protocol (MCP) servers. These servers provide the agent with capabilities to read/write the local filesystem, query databases (Supabase), and interact with GitHub.
[PROMPT_INJECTION]: The integration of MCP servers (such as Exa for web search and Firecrawl for web scraping) creates a surface for indirect prompt injection. Malicious instructions contained within web pages or files processed by these tools could potentially influence the agent's behavior.
Ingestion points: Data enters the agent's context through tools like filesystem, github, exa, and firecrawl as detailed in guides/07-mcp-and-ide-plugins.md.
Boundary markers: The skill does not provide specific instructions to use delimiters or ignore instructions within tool outputs to mitigate accidental obedience.
Capability inventory: All scripts in guides/ enable significant capabilities including filesystem access, database operations (Supabase), and GitHub repository management.
Sanitization: No explicit sanitization or validation of data retrieved from external tools is recommended within the guides.

ai-tools-platform-weapon