The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation identifies several external dependencies for MCP integration, including packages from trusted organizations such as Anthropic (@anthropic/stitch-mcp, @anthropic/figma-mcp) and well-known services like Fal.ai (@fal-ai/mcp) and OpenAI (openai-images-mcp).
[COMMAND_EXECUTION]: The skill provides instructions for configuring MCP servers using the npx command within the mcpServers configuration block. These are standard operational commands for the platform's tool integration system.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by design.
Ingestion points: The extract_design_context tool in references/mcp-design-tools.md accepts a source parameter which can be a URL or file path to an external design.
Boundary markers: The instructions do not explicitly mandate delimiters or 'ignore instructions' warnings for content fetched from these external URLs.
Capability inventory: The agent has the capability to read and comment on Figma files (via Figma MCP) and perform network-based generation tasks across various platforms.
Sanitization: No explicit sanitization or filtering logic is described for the content extracted from external design sources.
[SAFE]: The skill follows security best practices for credential management by instructing users to store sensitive API keys in environment variables or configuration files (~/.claude/mcp_servers.json) rather than hardcoding them in scripts.

ai-design-generation