Course Objective & Scope
Many agent failures come not from the model but from poorly designed tools: vague descriptions, too many parameters, cryptic error messages or overly broad permissions. This course covers the design principles for turning existing internal APIs into tools agents can use safely.
Tool contracts (input/output schema, pre- and postconditions, side effects, idempotency), pagination and response size control, versioning and permission scopes are practised hands-on. Participants turn one of the organisation's internal APIs into a tool set measured with evals and produce a design checklist.
Who Should Attend?
- Backend developers providing tools and integrations for agents
- AI engineers building agent applications
- API and platform teams responsible for exposing internal APIs to agents
- Prerequisite: familiarity with REST API design and JSON Schema
Key Takeaways
- Write tool names and descriptions an LLM can select and call correctly
- Define inputs and outputs through explicit tool contracts
- Design error messages that let the agent correct itself
- Apply idempotency and permission scopes to tools with side effects
- Measure and improve tool selection accuracy with an eval set
Curriculum
Tools from the Agent's Perspective
- Factors that influence the model's tool choice
- The difference between an API endpoint and an agent tool
- The principle of fewer, well-designed tools
Naming, Descriptions and Schemas
- Writing descriptions, examples and usage limits
- JSON Schema parameters, enums and required fields
- Strict mode and structured output
Tool Contracts
- Pre- and postconditions, side effects and idempotency
- Output format, pagination and response size control
- Versioning and backward compatibility
Error Design and Security
- Error messages the agent can act on
- Permission scopes, read/write separation and tools requiring approval
Measuring the Tool Set
- An eval set for tool selection and parameter accuracy
- Measuring the impact of description and schema iterations
- A design checklist for the organisation's API
Other Courses in This Category
-
Hands-on technical course on building production-ready agent systems with LangGraph: StateGraph, conditional edges, checkpointers, interrupt-based human-in-the-loop, subgraphs and the Send API.
- Live Course
- 6 modules
-
Course on building tool-based agents with the OpenAI Agents SDK and Responses API: function tools, hosted tools such as web search and file search, handoffs, guardrails, sessions and built-in tracing.
- Live Course
- 5 modules
-
Hands-on course on building Model Context Protocol servers that expose tools, resources and prompts, and clients that consume them: stdio and Streamable HTTP, OAuth-based authorisation, security and testing.
- Live Course
- 5 modules
-
Designing multi-agent systems with supervisor, hierarchical, orchestrator-worker and swarm patterns: context isolation, task routing, cost control and a comparison of LangGraph, CrewAI and Microsoft Agent Framework.
- Live Course
- 5 modules
-
Hands-on training on designing working, episodic, semantic and procedural memory layers for AI agents using context summarisation, vector and graph memory stores, memory consolidation and retention policies.
- Live Course
- 5 modules
-
Technical course on safely executing LLM- and agent-generated code using gVisor, Firecracker microVMs, container isolation, network/filesystem restrictions and resource quotas.
- Live Course
- 5 modules
Get in touch
