Programmatic interfaces, Model Context Protocol (MCP) servers, and machine-readable specifications for autonomous AI agents and engineering teams.
Oh my AI provides an open sandbox environment for developer agents and automated testing. You can exercise all intake and discovery endpoints without production side effects.
https://ohmyai.nl/apiX-Environment: sandbox or bearer token omai_sandbox_demoIdempotency-Key: <uuid> on write requestscurl -X POST https://ohmyai.nl/api/intake \
-H "Content-Type: application/json" \
-H "Idempotency-Key: $(uuidgen)" \
-H "X-Environment: sandbox" \
-d '{
"name": "Autonomous Agent",
"email": "[email protected]",
"process": "Automating customer intake and workflow qualification",
"timeline": "2-4 weeks",
"budget": "€5.000 - €15.000"
}'
AI agents and code generation tools can inspect and consume our formal contracts across all standard discovery surfaces:
JSON-RPC 2.0 tool endpoints for Claude, ChatGPT, and custom agents.
/.well-known/mcpDiscovery catalog adhering to the agentskills.io v0.2.0 schema.
/.well-known/agent-skills/index.jsonAgentic Resource Discovery v0.91 catalog with trust manifests.
/.well-known/ard.jsonOur production MCP server runs on Streamable HTTP transport and provides 4 native tools for LLM agent integration:
get_services — List production automation capabilities, deliverables, and SLAs.estimate_timeline — Calculate realistic timeline ranges based on workflow complexity.submit_inquiry — Programmatically submit scoped requests for engineering review.check_readiness — Run an audit check on workflow automation prerequisites.{
"mcpServers": {
"oh-my-ai": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-fetch", "https://ohmyai.nl/api/mcp"]
}
}
}
{"error": {"code": "...", "message": "...", "hint": "..."}}.RateLimit-Limit, RateLimit-Remaining, and RateLimit-Reset are included on every response.Accept: text/markdown to receive lightweight markdown versions of key pages.Access-Control-Allow-Origin: *).