a2a-agent-contract
CommunityStandardize Child Agent integration via A2A contracts.
System Documentation
What problem does it solve?
It prevents brittle, hard-coded Child Agent integrations by defining a universal A2A contract that Orchestrator, Registry, and ProtocolConverter can rely on, ensuring consistent discovery, health management, streaming behavior, and artifact normalization.
Core Features & Use Cases
- Defines Orchestrator ↔ Child Agent A2A契约: standardizes AgentCard, health check, sendSubscribe task contract, streaming lifecycle semantics, and artifact output as ArtifactDraft.
- Enforces capability-driven selection: requires Orchestrator/Planner to use AgentCard.skills and outputModes instead of agentName-based capability guessing.
- Guarantees safe event boundaries: restricts Child Agents from emitting AG-UI events or frontend tool calls; ProtocolConverter handles A2A→AG-UI mapping.
- Supports multi-agent expansion: makes the contract generic enough for 2+ agents, with optional future profiles (v1.0→post-v1.0) without locking specific agent names.
- Hardens security and fallback behavior: specifies health normalization, retryable failure handling, and anti-secret/anti-exfiltration rules.
Quick Start
Use this contract when onboarding or modifying any Child Agent by updating its AgentCard, /health, and /a2a/tasks/sendSubscribe streaming + ArtifactDraft outputs so Orchestrator and Registry can discover it and safely map its results to AG-UI.
Dependency Matrix
Required Modules
None requiredComponents
💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: a2a-agent-contract Download link: https://github.com/zxc1a1a1/Multi_Agent-AgentHub/archive/main.zip#a2a-agent-contract Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.