FutureStack Developer Platform
Access verified metadata, pricing models, and community reviews for 500+ AI software tools. Designed specifically for autonomous AI agents, LLM function calling, and developer workflows.
Model Context Protocol (MCP) Server
Connect Claude Desktop, Cursor, Windsurf, or custom agent runtimes to FutureStack using our standard Streamable HTTP & SSE MCP endpoint. No authentication token required for discovery and read operations.
{
"mcpServers": {
"futurestack": {
"url": "https://www.usefuturestack.com/api/mcp"
}
}
}Search 500+ AI tools by keyword, category, audience role, or pricing tier.
Retrieve full feature matrix, pricing plans, and verified ratings by tool slug.
Get the top community upvoted tools in real-time.
Retrieve today's featured community spotlight AI tool.
REST API Quickstart
Query the directory directly over HTTPS. All responses are served in typed JSON with CORS enabled for browser and serverless consumption.
curl -s "https://www.usefuturestack.com/api/tools?search=coding&pricing=Free&limit=5"import requests
res = requests.get(
"https://www.usefuturestack.com/api/tools",
params={"search": "video", "role": "Developer", "sort": "popular"}
)
data = res.json()
for tool in data.get("tools", []):
print(f"{tool['name']} ({tool['pricing_badge']}) - {tool['short_description']}")const res = await fetch('https://www.usefuturestack.com/api/tools?category=Coding&limit=10');
const { tools, total } = await res.json();
console.log(`Found ${total} tools:`, tools.map(t => t.name));Markdown Content Negotiation
FutureStack is compliant with the acceptmarkdown.com protocol. Pass the Accept: text/markdown header with any page request to receive clean Markdown formatted specifically for LLM token efficiency.
curl -H "Accept: text/markdown" https://www.usefuturestack.com/tools/cursorAPI Endpoints Reference
Full details and parameter schemas are specified in the OpenAPI 3.1 definition.
Search, paginate, and filter tools by category, role, and pricing.
Get tool profile, pricing plans, and verified review summaries.
Returns structured Markdown representation for LLM ingestion.
List top 10 trending AI tools.
Model Context Protocol JSON-RPC 2.0 tool execution endpoint.
Rate Limits & Policies
- Public Tier: 100 requests per minute per IP address. No API key required for standard discovery queries.
- Caching: GET responses include
Cache-Control: public, s-maxage=3600andVary: Accept, Accept-Encoding. - Error Format: All API errors return structured JSON with
error.code,error.message, anderror.hint.