Model Online / On-Site
Duration
Modules 5 modules

Course Objective & Scope

Stuffing dozens of business policies, hundreds of API tools and lengthy procedural manuals into a single system prompt inflates token cost and degrades tool-selection accuracy. The Agent Skills approach packages domain knowledge, workflow instructions, templates and helper scripts into modular capabilities loaded on demand.

This course covers skill metadata and trigger descriptions, progressive context disclosure, pairing deterministic scripts with LLM reasoning, combining MCP tools with skills and managing versioned skill registries. Participants package real internal standard operating procedures into portable, testable agent skills.

Who Should Attend?

  • AI engineers building broad-scope enterprise agents and coding or operations assistants
  • Platform engineers providing shared agent capabilities across teams
  • Technical leads and solution architects encoding domain workflows for agents
  • Prerequisite: familiarity with prompt engineering, tool calling and basic agent architecture

Key Takeaways

  • Distinguish low-level atomic tools from high-level domain skills and capabilities
  • Apply progressive disclosure to keep initial context lean and load detailed instructions only when triggered
  • Author modular skill bundles combining instructions, reference docs, templates and verification scripts
  • Implement dynamic skill and tool retrieval over large enterprise tool catalogues
  • Establish an internal skill registry with versioning, permission boundaries and evaluation tests

Curriculum

From Tools to Skills: Capability Design Principles
  • Why raw API endpoints alone do not constitute an agent capability
  • Combining tools, domain rules, workflows and verification into a cohesive skill
  • The Agent Skills standard and SKILL.md package architecture
Progressive Disclosure and Context Economy
  • Tier 1 (metadata/triggers), Tier 2 (core instructions) and Tier 3 (reference files and scripts)
  • Runtime skill discovery and on-demand activation
  • Semantic tool and skill routing across large enterprise catalogues
Skill Authoring and Deterministic Helpers
  • Writing unambiguous trigger descriptions, step-by-step procedures and examples
  • Bundling pre-tested scripts and templates instead of regenerating code from scratch
  • Pairing MCP servers (connectivity) with Agent Skills (procedural know-how)
Capability Composition, Permissions and Governance
  • Composing multiple skills within a single task without instruction conflicts
  • Skill-scoped tool allowlists and security boundaries
  • Versioning, eval suites and managing an internal enterprise skill registry
Workshop: Building an Enterprise Skill Library
  • Packaging a real internal operating procedure into a reusable skill bundle
  • Measuring trigger precision, token savings and task completion quality

Other Courses in This Category

  • Agent Interoperability and A2A (Agent2Agent) Protocols

    Course on cross-framework and cross-vendor agent collaboration using the A2A (Agent2Agent) protocol: Agent Cards, task lifecycle, streaming updates, enterprise authentication and MCP–A2A complementarity.

    AI Agent Engineering Corporate
    Live Course
    5 modules
    View
  • Human Approval and Escalation: Human-in-the-Loop Agent Workflows

    Training on pausing agents for human approval on high-risk tool calls and low-confidence decisions, editing tool arguments, routing asynchronous escalations via Slack/Teams/web UIs and maintaining audit trails.

    AI Agent Engineering Corporate
    Live Course
    5 modules
    View
  • ReAct Pattern for LLM Agents: Reasoning and Acting

    Hands-on training for developers on building ReAct agents from scratch: the Thought–Action–Observation loop, native tool calling, stop conditions, iteration limits and failure analysis.

    AI Agent Engineering Corporate
    Live Course
    5 modules
    View
  • Plan-and-Execute Agents: Planning, Execution and Replanning

    Technical course on agents that split multi-step tasks into planner, executor and replanner roles, covering ReWOO and LLMCompiler-style parallel task graphs, plan validation and cost optimisation.

    AI Agent Engineering Corporate
    Live Course
    5 modules
    View
  • State and Workflow Management for AI Agents

    Course for teams building long-running agents: state schemas, reducers, checkpoints, thread management and state-machine workflows, illustrated with LangGraph and workflow engine examples.

    AI Agent Engineering Corporate
    Live Course
    5 modules
    View
  • Tool Design and Tool Contracts for LLM Agents

    Hands-on course on designing tools agents can use reliably: naming, descriptions, JSON Schema parameters, error messages, tool contracts, versioning and permission scopes.

    AI Agent Engineering Corporate
    Live Course
    5 modules
    View

Get in touch