# Agentic DX: Bringing IDP into your IDE

## Executive summary

This session details how a team tackled low developer adoption of an Internal Developer Platform (IDP) by implementing 'Agentic DX'—bringing agent functionality directly into the IDE. The solution centers on a Master Agent architecture that orchestrates multiple specialized agents using A2A protocols, ensuring secure and reliable context management while mitigating issues like hallucination snowballing and redundant development efforts.

## Key takeaways

- Addressing Low IDP Adoption: To overcome the 'chicken and egg' problem of low developer adoption (e.g., 22% usage), the team pivoted to a 'Bring Your Own Agent' strategy, integrating agentic capabilities directly into the IDE where developers naturally work.
- Master Agent Architecture: The Master Agent acts as a single entry point and orchestrator, deciding which specialized agent should handle a user query. This approach prevents the complexity of naive canonical meshes where every agent can see every other agent.
- Context Hygiene via Decomposition: To combat context pollution and improve reliability, the system was decomposed into four distinct loops: Perception (planner), Decision (executor), Action (tool caller), and Memory. This structured approach ensures each component has a defined responsibility.
- Security and Communication Protocols: The platform mandates 'Zero Trust Execution' and requires human approval for every tool call or agent interaction. All agent-to-agent communication must adhere to the A2A protocol.

## Technical details

- Agentic Architecture: The system utilizes a Master Agent connected via an MCP endpoint (e.g., VS Code). The flow involves the user query triggering a pool call, followed by context enrichment from the agent registry, and finally execution by the appropriate specialized agent (e.g., FinOps agent) using A2A calls.
- Protocol Mapping: The architecture requires careful mapping between different protocols, specifically detailing the conversion of a consent URL received via an A2W (Agent to Workflow) request state into an MCP elicitation for human approval.
- Context Management: To reduce context overhead, the implementation of a 'tool search tool' was critical, achieving up to an 80% reduction in context space by dynamically performing semantic searches on required tools and skills.

## Practical implications

- When building internal developer platforms (IDPs), consider integrating agentic capabilities directly into the IDE to maximize adoption and solve the 'chicken and egg' problem.
- Adopt a Master Agent pattern rather than relying on naive, fully connected agent meshes to manage complexity and prevent hallucination snowballing.
- Implement strict context hygiene by decomposing the LLM process into distinct stages (Perception, Decision, Action, Memory) for reliability.

## Topics

Agentic AI, Internal Developer Platform (IDP), Developer Experience (DX), Master Agent Architecture, A2A Protocol, Context Management, Agentic AI Foundation (AAIF), A2W and C2C Protocols

Source: https://www.youtube.com/watch?v=u3ocH-qdyi4