Subagent Prompt Construction

Construct compact prompts by using a lambda-contract structure within a 150-line cap, defining explicit dependencies, organizing into three layers (orchestration, tool integration, skill references), and validating with automated tooling to ensure agent composition and MCP tool integration work correctly.

Use orchestrated prompt patterns that compose multiple agents and MCP tools together. The Skill provides reference patterns and integration templates that show how to structure agent composition, define tool contracts, and layer skill references so agents coordinate without exceeding the 150-line constraint.

Validation scripts extract patterns from your prompts and check against functional requirements: strict 150-line limit, explicit dependency declarations, three-layer architecture compliance, and lambda contract correctness. Automated tooling flags violations before deployment.

Yes. The Skill supports embedding skill references and symbolic logic as validated components within subagent prompts. Reference content and validation patterns ensure symbolic logic integrates correctly into the three-layer architecture without violating the 150-line compact constraint.

Identify your MCP tools, list agent dependencies explicitly, define lambda contracts for each component interaction, and gather reference patterns for your use case (orchestration, analysis, enhancement, or validation). The Skill template guides setup for these scenarios.