skills.rest
agentic-community
agentic-communityOFFICIAL
Explore all skills by this author →
GitHub

generate-agent-card

Generate A2A agent card JSON by analyzing agent source code in folders or GitHub URLs.

892|225|Updated May 29, 2025
Search Tags:#Validation#A2A#Protocol#agent registry#agent card#JSON generation#Code Analysis
One-click install
npx skills add https://github.com/agentic-community/mcp-gateway-registry --skill generate-agent-card
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: generate-agent-card
Source: https://github.com/agentic-community/mcp-gateway-registry/tree/main/.claude/skills/generate-agent-card
Command: npx skills add https://github.com/agentic-community/mcp-gateway-registry --skill generate-agent-card

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Manually writing a spec-compliant A2A (Agent-to-Agent) agent card requires reading through an entire agent codebase to identify its name, skills, tools, protocol bindings, and authentication mechanisms, which is tedious and error-prone for multi-file agent projects.

Core Features & Use Cases

  • Deep Code Analysis: Explores entire folder structures including entrypoints, tool definitions, prompts, MCP server connections, and config files to detect agent capabilities.
  • Spec-Compliant Output: Generates agent card JSON following the official A2A specification with required fields like supportedInterfaces, capabilities, skills, and securitySchemes.
  • Built-in Validation: Runs a Python validation script that checks JSON format, required fields, interface bindings, and skill entries before reporting success.
  • Use Case: Point the skill at a Strands or BedrockAgentCore agent repository and receive a validated agent card JSON with detected skills, protocol binding (JSONRPC or HTTP+JSON), and auth schemes like SigV4 or Cognito JWT.

Quick Start

Generate an A2A agent card for the agent source code in the folder ./my-agent and save the validated JSON next to it.

Frequently Asked Questions about generate-agent-card

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate an A2A agent card from source code?▼

Provide a local folder path or GitHub URL containing the agent code. The skill explores all Python files, configs, and README files, detects the agent name, skills, protocol, and auth, then writes a validated agent card JSON into the agent's folder.

What fields are required in an A2A agent card?▼

Required fields are name, description, version, supportedInterfaces, capabilities, defaultInputModes, defaultOutputModes, and skills. Each skill entry needs id, name, description, and tags, and each interface needs url, protocolBinding, and protocolVersion.

Does the agent card generator support MCP tools as skills?▼

Yes, it detects MCP server connections in the agent code, such as MCPClient configs and tool imports, and lists those MCP tools as skills in the generated card by reading the MCP server configurations for tool names and descriptions.

How is the protocol binding detected for the agent card?▼

A2A agents using a2a_server or A2AServer get JSONRPC binding, while HTTP agents using BedrockAgentCoreApp or REST endpoints get HTTP+JSON. If detection is ambiguous, the skill defaults to JSONRPC.

Why does agent card validation fail after generation?▼

Validation fails when required fields are missing, protocolBinding is not JSONRPC, GRPC, or HTTP+JSON, or skills lack id, name, description, or tags. The skill reports each error, fixes the JSON, and re-validates until all checks pass.

Related Skills

Explore More Software Engineering Skills→
lando-labslando-labs

manage-sources

Manage Git-based and local AI agent capability sources with version control.

Official
Advanced
Heretek-AIHeretek-AI

a2a-agent-register

Registers collective agents with the LiteLLM A2A gateway via a configurable workflow.

Official
Advanced
LPDigital-AgentLPDigital-Agent

backend-architect

Design serverless FastAPI backends with AWS Lambda, DynamoDB, S3, and Cognito.

Official
Advanced
OrcaQubitsOrcaQubits
35

a2a-mcp-bridge

Bridges Agent-to-Agent protocol with Model Context Protocol for hybrid systems.

Official
Advanced