Atlas › 01 Agent Skills for Browser Automation › Skills vs MCP: Architectural Comparison☰ Read as one page
Skills vs MCP: Architectural Comparison
7.1The Fundamental DifferenceMCP connects Claude to external capabilities (tools, data, APIs). Skills teach Claude how to use those capabilities for specific domains.7.2How Each Approach WorksWhat's loaded per API call: - 15-25 tool schemas (~5,000-12,500 tokens) - Accessibility tree per page (~2,000-10,000 tokens), streamed…7.3Detailed ComparisonThe mechanism (schemas every turn, accessibility trees inline) can produce much larger per-step ratios on paper — but the honest, measured…7.4Decision Matrix: When to Use What1. Your agent has filesystem access — Claude Code, Cursor, CI runners. The disk is the context-overflow valve; this is the whole trick.7.5Microsoft's Own AssessmentThis used to be the part of the course where we argued the token math from first principles. As of 2026, we can simply cite the vendor…7.6Architecture Diagram: Hybrid ApproachIn this hybrid model: - playwright-cli handles interaction commands (open, goto, click, fill, snapshot, screenshot) — cheap, disk-first…7.7Interview Talking Point"We evaluated both MCP and CLI-skill approaches for browser automation. The key trade-off is where page state goes. MCP streams tool…