Godle
Free AI prompt orchestration with 185 expert roles, 1,741 templates, and 4 MCP tools — no API key required
README
GODLE SDK
Free AI prompt orchestration SDK with 185 expert roles, 1,741 prompt templates, and agent protocol support (MCP + A2A).
No API key required. Works in browser and Node.js. Zero dependencies.
What is GODLE?
GODLE is a free, open AI prompt library and orchestration standard. It provides structured, expert-level prompt templates for 185 professional roles across 24 categories — from software engineering and data science to marketing, finance, legal, HR, and more.
Every template includes input/output schemas, eval rubrics, and model routing hints. The SDK connects to a two-layer architecture:
- Layer 1 (free, no auth): Static JSON API with all roles, templates, workflows, and eval rubrics
- Layer 2 (API key): Stateful edge layer with sessions, task execution, streaming, and eval-driven retries
Quick start
Browser
<script src="https://godle.app/api/v3/godle-sdk.js"></script>
<script>
// List all roles
const roles = await GODLE.listRoles();
console.log(roles); // 185 roles
// Get a specific role with all templates
const swe = await GODLE.getRole('software-engineering');
console.log(swe.templates); // 10 prompt templates
// Generate an expert prompt
const prompt = await GODLE.generatePrompt('software-engineering', 'system_design', {
input: 'Design a real-time notification system for 10M users'
});
console.log(prompt); // Full structured prompt ready for any LLM
</script>
Node.js
// Using the bundle
const GODLE = require('./sdk/godle-sdk.bundle.js');
// Or import individual modules
const { listRoles, getRole, generatePrompt } = require('./sdk/godle-sdk-core.js');
const roles = await listRoles();
const prompt = await generatePrompt('data-science', 'exploratory_data_analysis', {
input: 'Analyze customer churn patterns in our SaaS product'
});
SDK modules
| Module | What it does |
|---|---|
| godle-sdk-core.js | listRoles(), getRole(), generatePrompt(), evaluateOutput() — core read-only API |
| godle-sdk-adapters.js | LLM provider adapters for Anthropic and OpenAI with model routing |
| godle-sdk-workflows.js | DAG workflow execution with parallel steps, eval-driven retries, token budgets |
| godle-sdk-sessions.js | Persistent sessions with role-scoped memory and task history |
| godle-sdk-streaming.js | SSE streaming client with heartbeat monitoring and auto-reconnect |
| godle-sdk-mcp.js | MCP client and server — JSON-RPC 2.0 tool protocol |
| godle-sdk-a2a.js | Google A2A client and server — agent-to-agent task dispatch |
| godle-sdk.bundle.js | All 7 modules bundled as UMD (~128KB) |
Available roles (185)
Roles span 24 professional categories:
Engineering & IT: Software Engineering, Frontend Development, Backend Development, DevOps & SRE, ML Engineering, Data Engineering, QA Engineering, Security Engineering, Cloud Architecture, Solutions Architecture, Technical Writing
Data & Analytics: Data Science, Business Intelligence, Data Analytics, Research Science
Product & Design: Product Management, UX Design, UI Design, Product Operations
Marketing: Content Marketing, Growth Marketing, SEO, Social Media, Brand Marketing, Email Marketing, Product Marketing
Sales: Enterprise Sales, Mid-Market Sales, SDR/BDR, Sales Operations, Account Management
Finance: FP&A, Accounting, Treasury, Tax, Audit, Investor Relations
HR & People: HR Business Partnering, Recruiting, L&D, Compensation & Benefits, DEI, People Analytics
Legal: Corporate Law, Compliance, Contract Management, Legal Operations, Privacy/Data Protection
Operations: Supply Chain, Procurement, Facilities, Business Operations
Strategy: Corporate Strategy, M&A, Management Consulting, Business Development
Customer: Customer Success, Customer Support, Community Management
...and more. See the full role list or call GODLE.listRoles().
API endpoints (free, no auth)
All Layer 1 endpoints return JSON with CORS enabled. No API key needed.
GET https://godle.app/api/v3/roles.json # All 185 roles
GET https://godle.app/api/v3/roles/{slug}.json # Full role + templates
GET https://godle.app/api/v3/categories.json # 24 categories
GET https://godle.app/api/v3/workflows.json # 12 multi-step workflows
GET https://godle.app/api/v3/evals.json # 1,741 eval rubrics
GET https://godle.app/api/v3/team-packs.json # 8 agent team compositions
GET https://godle.app/api/v3/capabilities.json # Searchable capability index
GET https://godle.app/api/v3/index.json # API manifest
MCP integration
GODLE exposes 4 MCP tools that any MCP-compatible AI agent can call:
{
"tools": [
"godle_list_roles",
"godle_match_capability",
"godle_execute_task",
"godle_compose_workflow"
]
}
MCP manifest: https://godle.app/.well-known/mcp.json
// Use as MCP client
const client = GODLE.mcp.client('https://godle.app');
const tools = await client.listTools();
const result = await client.callTool('godle_list_roles', { categoryId: 'engineering-it' });
// Expose as MCP server
const server = GODLE.mcp.server();
server.addTool('generate_prompt', { /* schema */ }, async (params) => {
return GODLE.generatePrompt(params.roleId, params.templateKey, params);
});
A2A integration
GODLE implements Google's Agent-to-Agent protocol for multi-agent orchestration:
A2A agent card: https://godle.app/.well-known/agent.json
// Send a task to GODLE via A2A
const client = GODLE.a2a.client('https://godle.app');
const task = await client.sendTask({
skill: 'software-engineering',
message: 'Design a microservices architecture for an e-commerce platform'
});
// Re-expose GODLE as your own A2A agent
const server = GODLE.a2a.server({ name: 'my-agent', skills: ['code-review'] });
const card = server.generateAgentCard();
Workflows
12 pre-built multi-step workflows with DAG execution:
| Workflow | Steps | What it does |
|---|---|---|
| feature-development | 6 | PRD to code review pipeline |
| incident-response | 5 | Detect, triage, fix, postmortem |
| content-campaign | 7 | Strategy to distribution |
| hiring-pipeline | 6 | JD to offer letter |
| design-sprint | 5 | Research to prototype |
| data-pipeline-build | 5 | Schema to monitoring |
| compliance-audit | 6 | Scope to remediation |
| launch-checklist | 8 | Pre-launch to post-launch |
| sales-deal-cycle | 7 | Prospect to close |
| quarterly-planning | 5 | OKRs to review |
| onboarding-playbook | 5 | Pre-boarding to 90-day check |
| vendor-evaluation | 6 | Requirements to contract |
const run = await GODLE.run('feature-development', {
feature: 'Add real-time notifications',
context: 'B2B SaaS, 50k users, React frontend, Node backend'
});
Team packs
8 pre-configured agent team compositions:
| Pack | Roles | Use case |
|---|---|---|
| product-trio | PM, Designer, Engineer | Feature development |
| saas-mvp | 6 roles | Ship an MVP |
| content-engine | 4 roles | Content pipeline |
| growth-team | 5 roles | Growth experimentation |
| data-team | 4 roles | Data pipeline + analytics |
| enterprise-sales-team | 4 roles | Enterprise deal cycle |
| incident-response-team | 4 roles | Incident management |
| compliance-team | 4 roles | Compliance audit |
Examples
See the examples/ directory:
- examples/browser.html — Browser usage with CDN script tag
- examples/node.js — Node.js usage with require/import
- examples/mcp-client.js — MCP client integration
File structure
godle-sdk/
sdk/ # SDK source modules
godle-sdk-core.js # Core API (listRoles, getRole, generatePrompt)
godle-sdk-adapters.js # LLM provider adapters (Anthropic, OpenAI)
godle-sdk-workflows.js # DAG workflow execution engine
godle-sdk-sessions.js # Session management with memory
godle-sdk-streaming.js # SSE streaming client
godle-sdk-mcp.js # MCP client + server
godle-sdk-a2a.js # A2A client + server
godle-sdk.bundle.js # All-in-one UMD bundle (~128KB)
docs/
api-v3.md # API reference
migration-v2-to-v3.md # v2 to v3 migration guide
SPEC-AGENTIC-LAYER.md # Agentic architecture spec
manifests/
mcp.json # MCP tool manifest
agent.json # A2A agent card
ai-plugin.json # ChatGPT plugin manifest
openapi.json # OpenAPI 3.1 spec
examples/
browser.html # Browser quick start
node.js # Node.js quick start
mcp-client.js # MCP integration example
llms.txt # Compact LLM reference
llms-full.txt # Extended reference with template examples
LICENSE # MIT License
package.json
Links
- Website: https://godle.app
- AI prompt generator: https://godle.app/app
- Jobs (all 185 roles): https://godle.app/jobs
- API docs: https://godle.app/api
- API playground: https://godle.app/app/playground
- Blog: https://godle.app/blog
- MCP manifest: https://godle.app/.well-known/mcp.json
- A2A agent card: https://godle.app/.well-known/agent.json
- OpenAPI spec: https://godle.app/api/v3/openapi.json
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。