// chapter by chapter · 19 chapters
AI Agent Studio guide 26D
Source: OracleHow do I use AI Agent Studio? 26D ↗
New in 26D
- OverviewWhat AI Agent Studio is and its main parts.
- Access Requirements for AI Agent StudioRoles and privileges people need to use it.
- Before You BeginSetup to finish before building.
- Use CasesPatterns and examples of what you can build.
- Expressions in AI Agent StudioPass values between steps.
- AI AgentsCreate and set up agents.
- Use Workflow BuilderBuild step-by-step workflow agents.
- Use Applications BuilderPut agents together into agentic apps.
- Edit Agents, Workflows and Agentic ApplicationsChange drafts and Oracle-delivered items safely.
- Debug and Test AI AgentsTry agents and find problems before going live.
- AI Unit Budgeting and ConsumptionNewSet budgets and warnings for AI usage.
- Monitor and Evaluate AI AgentsWatch how agents perform and measure quality.
- Promote Published AgentsMove finished agents from test to production.
- Use AI Studio Command Line Interface (CLI)Build agents as files from the command line.
- Integrate Applications and Workflows with Continuous Integration and Continuous Delivery (CI/CD)NewRelease apps and workflows through Git and pipelines.
- Use Agentic Testing and Lifecycle Automation Suite (ATLAS)NewRecord and replay workflows for repeatable tests.
- Allow External Access to Fusion Applications AgentsLet outside applications call Fusion agents.
- Collaborate with AI Agents Across Platforms Using Agent2Agent (A2A) ProtocolWork with agents on other platforms.
- AI Agent Studio FAQsAnswers to common questions.