Chapter AG10-C01 · Agents, tools, memory, and bounded execution
Architecture and tool boundaries
Choose the simplest adequate execution design; Validate arguments and separate read from write capabilities.
Units in this chapter
- AG10-U01 Workflows versus autonomous loops Choose the simplest adequate execution design
- AG10-U02 Tool schemas and authority Validate arguments and separate read from write capabilities
Reading that connects these units
Reading
- Building effective agents Compare deterministic workflows, tool-augmented calls, and autonomous loops.About this reference (W18)
- AI Engineering Application patterns, evaluation, and system-level tradeoffs; select AI Engineering on the author page.About this reference (B09)
- Model Context Protocol documentation Protocol roles, client/server boundaries, tools, and authorization concepts.About this reference (W24)
- MCP Python SDK Inspect the minimal client/server and tool examples before adding permissions.About this reference (G12)
- OWASP GenAI Security Project Prompt injection, sensitive information, excessive agency, and application threat models.About this reference (W26)
Selected, not required. Reviewed 2026-11-17.