Når en dansk tjeneste mangler et officielt API eller en færdig MCP-server, gør disse fundamentale værktøjer det muligt for udviklere og agenter at bygge CLI'er, generere MCP-servere fra OpenAPI-specifikationer og køre browserautomatisering.
Tools and microframeworks that instantly convert OpenAPI specs, Swagger docs, or raw APIs into production-ready Model Context Protocol (MCP) servers and CLI binaries.
Vision-driven web agents and scalable headless Chrome infrastructure to automate clicks, forms, and data collection on legacy portals without APIs.
Engines that crawl complex websites, bypass anti-bot challenges, and convert unstructured HTML into pristine, token-efficient Markdown for agent context.
Isolated micro-VMs and execution containers for AI agents to safely run Python scripts, shell commands, and automated data processing pipelines.
Instant MCP & CLI generator from OpenAPI, Swagger & APIs
PrintingPress converts OpenAPI and Swagger specifications into fully-typed Model Context Protocol (MCP) servers and CLI tools in seconds.
High-level Python & TypeScript framework for building MCP servers
FastMCP provides clean decorators and microframework patterns to write production-ready MCP tools in under 10 lines of code.
Turn entire websites and portals into clean, LLM-ready Markdown
Firecrawl crawls, handles JavaScript rendering, bypasses anti-bot barriers, and outputs clean Markdown structured for AI agent ingestion.
Autonomous open-source AI agent browser automation engine
Browser-use enables LLMs to interact directly with web browsers — clicking buttons, filling forms, and navigating complex workflows like a human.
AI-powered web automation and structured data extraction SDK
Built on Playwright by Browserbase, Stagehand provides high-level `act()`, `extract()`, and `observe()` primitives for resilient web automation.
Scalable headless Chrome fleet for Playwright and Puppeteer agents
Browserless provides enterprise-grade, cloud or self-hosted headless browser infrastructure designed to run high-concurrency agent workflows.
W3C-standardforslag til eksponering af strukturerede AI-værktøjer direkte på websider
WebMCP gør det muligt for webapplikationer at eksponere typede JavaScript-funktioner og annoterede HTML-formularer direkte til AI-agenter i browseren.
Open-source, blazing fast asynchronous LLM web crawler
Crawl4AI is an open-source, speed-optimized asynchronous web crawler that generates structured Markdown and JSON graphs for AI models.
Secure cloud sandboxes for autonomous AI agent code execution
E2B provides isolated, sub-second cloud sandboxes where AI agents can execute arbitrary Python, JavaScript, and shell code safely.
Zero-setup URL-to-Markdown proxy for any web page
Prepend `r.jina.ai/` to any URL to instantly convert the page into clean, LLM-ready markdown with zero authentication or setup.
Open-source development environment & agent workspace manager
Daytona provides standardized, containerized agent workspaces and development environments across local and cloud infrastructure.
Convert browser network captures (HAR) and API traffic into MCP tools
Har2MCP inspects recorded browser network activity (.har files) and automatically reverse-engineers clean Model Context Protocol tools.
Official Model Context Protocol terminal runner, inspector & tester
The standard developer CLI and interactive web inspector to test, debug, and execute MCP servers locally and remotely.
Model Context Protocol wrapper for Playwright browser automation
Enables LLMs and agent harnesses to control real browser sessions via standard MCP tool calls (navigate, click, type, screenshot).
Search engine & content extractor built exclusively for autonomous AI agents
Tavily is a purpose-built search engine designed for LLMs, delivering clean factual snippets and real-time web context with zero noise.
Developer-first headless browser platform with stealth, residential proxies & session recording
Browserbase is cloud headless browser infrastructure featuring built-in CAPTCHA solving, residential proxy rotation, and session video replays.
Structured data extraction & type-safe LLM function schemas
Instructor enforces strict Pydantic and Zod schemas on LLM outputs, turning unstructured web and API data into typed objects with automatic retries.
Open-source browser infrastructure built exclusively for AI agents
Steel provides production-grade headless browser sandboxes with built-in stealth, session persistence, and anti-bot mitigation.
Neural search engine & structured content extractor for AI agents
Exa is an embedding-powered web search engine designed for LLMs, delivering clean semantic snippets and structured page extraction.
AI-powered semantic query language for web scraping and automation
AgentQL uses natural language queries to locate web elements and extract structured data without brittle CSS selectors or XPath.
Type-safe prompt programming language & LLM function compiler
BAML is a domain-specific language that turns complex LLM prompts into strongly-typed functions with automatic client generation.
Proxy stdio MCP servers over SSE & WebSockets for remote AI agents
Supergateway bridges local stdio-based Model Context Protocol servers to Server-Sent Events (SSE) and WebSockets for remote cloud agents.
High-performance Rust web crawler & scraper generating LLM-ready markdown
Spider is an ultra-fast web crawler written in Rust that streams clean Markdown, structured JSON, and embeddings for AI models.
Open-source LLM observability, tracing & prompt management platform
Langfuse provides full-lifecycle tracing, evaluation, and monitoring for AI agents executing complex multi-step tool calls.
Convert any REST OpenAPI or Swagger spec into instant MCP tools
Dynamically translates OpenAPI 3.0/3.1 specifications into Model Context Protocol tools with automated parameter validation.
Automatiseret agentparatheds-scanner der tester webløsninger for MCP, A2A, WebMCP og Auth.md
Cloudflares automatiserede scanner, der evaluerer webløsninger inden for opdagelse, indholdstilgængelighed, bot-adgang, protokolstandarder og handel.
Sammensæt sikre AI-agenter ud fra udskiftelige plugins, værktøjer og runtimes
Et open-source, plugin-baseret agent-runtime fra DeepSeek AI med CLI, Web UI, Python SDK, værktøjsregister, persistente sessioner og sandkassestyring.
Hvis du har udviklet et open-source værktøj, crawler eller browserautomatiseringsflow, der hjælper agenter med at tilgå dansk infrastruktur, så tilføj det til AgentAccess.