arcade-mcp

Arcade MCP Server Framework: a Python framework/CLI for creating Model Context Protocol (MCP) servers with an opinionated project template, a runtime (stdio or HTTP transport), and support for tool-level requirements like secrets and OAuth-based auth injected at runtime.

Evaluated Mar 30, 2026 (22d ago)
Homepage ↗ Repo ↗ DevTools mcp model-context-protocol python tool-calling agent-integration framework oauth secrets
⚙ Agent Friendliness
56
/ 100
Can an agent use this?
🔒 Security
78
/ 100
Is it safe for agents?
⚡ Reliability
40
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
85
Documentation
70
Error Messages
0
Auth Simplicity
65
Rate Limits
20

🔒 Security

TLS Enforcement
80
Auth Strength
75
Scope Granularity
80
Dep. Hygiene
65
Secret Handling
85

Secrets and OAuth tokens are described as injected into a server-side Context so LLMs/MCP clients cannot see them. Tool-level auth uses scopes (fine-grained). TLS for HTTPS HTTP transport is implied by an HTTP endpoint and doc reference to /docs; exact enforcement is not verifiable from provided content. Dependency hygiene appears decent from pinned versions in the manifest, but no CVE/security audit data is provided.

⚡ Reliability

Uptime/SLA
0
Version Stability
65
Breaking Changes
55
Error Recovery
40
AF Security Reliability

Best When

You want to implement MCP tools in Python with consistent patterns for authentication, secrets, and deployment/transport to popular MCP clients.

Avoid When

You need a fully specified REST/OpenAPI service for web clients only, or you require a hosting-free, zero-ops managed offering.

Use Cases

  • Build custom MCP servers with Python quickly (generate a project template, implement @app.tool functions)
  • Expose agent tools over MCP for Claude Desktop/Cursor/VS Code via stdio or HTTP transport
  • Wrap third-party APIs (e.g., Reddit OAuth) as callable MCP tools with per-tool auth scopes
  • Manage sensitive credentials by requiring secrets that are injected into the tool context at runtime

Not For

  • A standalone hosted API service with managed endpoints (it’s a framework you run/deploy)
  • Replacing a general-purpose REST backend or database layer
  • Use where you cannot run code (it requires your server runtime)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: CLI-based login (arcade login) Tool-level OAuth auth injection via arcade_mcp_server.auth (e.g., Reddit(scopes=[...])) Tool-level secret injection via requires_secrets and Context.get_secret
OAuth: Yes Scopes: Yes

The README example shows tool-level auth configured with a provider object (Reddit with scopes) and secrets injected into a Context at runtime; clients/LLMs cannot directly view secrets/tokens.

Pricing

Free tier: No
Requires CC: No

No pricing information in provided content; appears to be open-source framework.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Tool results may include provider/API errors (e.g., httpx response.raise_for_status()). Agents may need to handle error strings/HTTP-related exceptions surfaced by the server.
  • HTTP transport example notes that tools requiring auth/secrets may require deployment (arcade deploy / Developer Dashboard) rather than local HTTP usage, which can surprise local testing workflows.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for arcade-mcp.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered