Which tools can your
agent actually trust?
Assay scores MCP servers, APIs, and SDKs across agent friendliness, security, and reliability — so agents and developers can choose the right tools.
8642 packages evaluated across 26403 cataloged · 131 categories · avg score 50.6
Top Rated
View all →Stripe API
The developer's payment platform with APIs for accepting payments, managing subscriptions, handling billing, and fraud prevention, with an official MCP server.
Cloudflare DNS API
Cloudflare's REST API for managing DNS records, zones, page rules, firewall rules, and other Cloudflare services programmatically across managed domains.
E2B Code Interpreter API
E2B provides secure, isolated cloud sandboxes for running AI-generated code. Each sandbox is a full Linux VM that spins up in ~150ms, letting agents execute Python, JavaScript, bash, and other code safely without risk to host systems. Purpose-built for AI agents — supports file uploads, persistent processes, long-running sessions (up to 24 hours), and real-time stdout/stderr streaming.
HashiCorp Vault API
Secrets management platform API for storing, accessing, and rotating secrets (API keys, passwords, certificates, database credentials). Provides dynamic secrets, encryption-as-a-service, and fine-grained access control.
Neon
Neon is a serverless PostgreSQL platform that separates compute from storage to enable instant database branching, scale-to-zero, and autoscaling. Each Neon branch is a full copy-on-write snapshot of the database, enabling safe schema migrations, isolated test environments, and per-PR previews with zero additional storage cost until data diverges. The Management API (REST) controls infrastructure: creating projects, branches, compute endpoints, and roles. The official Neon MCP server (25+ tools) gives AI agents direct control over the full Neon platform — running SQL, managing migrations on temporary branches, analyzing slow queries, and provisioning databases — making Neon one of the most agent-ready databases available.
Adyen API
Enterprise global payment platform providing end-to-end acquiring, payment processing, risk management, and financial services for large-scale merchants across online, in-store, and mobile channels.
Recently Evaluated
Browse all →zyz-mcp-server
Provides an MCP (Model Context Protocol) server integration package (zyz-mcp-server), intended to expose tools/resources to MCP-capable agents.
zvec-mcp-server
Provides an MCP (Model Context Protocol) server exposing tools to manage a Zvec embedded vector database (collections, CRUD for documents, vector similarity search, index management) and optionally generate embeddings via OpenAI using environment-configured API credentials.
zstack-mcp-server
An MCP (Model Context Protocol) server implementation for “zstack” that exposes ZStack-related functionality to MCP-compatible AI agents/tools.
zsb-mcp-server
An MCP (Model Context Protocol) server implementation intended to expose tools/actions to MCP-compatible AI agents under the package name zsb-mcp-server.
zq145-mcp-server-ex0804
A custom MCP server package ("zq145-mcp-server-ex0804") intended to expose tool(s) to an MCP-compatible agent runtime. The provided prompt contains only the package name; no README, API/tool list, or configuration details were included to verify functionality.
zmy-mcp-server
An MCP (Model Context Protocol) server package intended to be used by LLM agents to access some external capability(s). The provided prompt includes only the package name and no README/repo content, so the specific tools/endpoints and behavior cannot be verified.
Categories
All categories →Query Assay at runtime
Your agent asks "what's the best payments tool?" — Assay answers in one database read. No LLM at query time. No hallucinated package names.
- ✓ REST API — JSON, OpenAPI spec available
- ✓ MCP Server — mount as a native agent tool
- ✓ Scores, gotchas, alternatives — one call
- ✓ Free tier: 100 queries/day, no account needed
{
"packages": [{
"id": "supabase",
"af_score": 78,
"has_mcp_server": true,
"gotchas": ["RLS must be..."],
"best_when": "Postgres + auth..."
}]
}
Stay in the loop
Get weekly updates on new evaluations, score changes, and ecosystem insights.
No spam. Unsubscribe anytime.
How Assay works
Discover
Agents continuously discover MCP servers, APIs, and SDKs across GitHub, npm, PyPI, and vendor docs.
Evaluate
Each package is scored across three dimensions — Agent Friendliness, Security, and Reliability — with detailed sub-scores for transparency.
Query
Agents (and developers) query the directory at runtime to select the right tool for the job — via REST API or MCP server.