mcp-kit

mcp-kit is a Go-based toolkit that provides an MCP server and MCP client implementation (using github.com/shaharia-lab/goai/mcp), plus an HTTP API server layer that exposes REST endpoints and orchestrates requests between a frontend, the MCP client, and LLM providers. It also includes monitoring/observability via Prometheus/Grafana/Loki/Promtail and provides an OpenAPI schema file (openapi.yaml) for the API server.

Evaluated Apr 04, 2026 (16d ago)
Repo ↗ Ai Ml mcp model-context-protocol go sse llm-integration http-api tooling observability self-hosted
⚙ Agent Friendliness
48
/ 100
Can an agent use this?
🔒 Security
57
/ 100
Is it safe for agents?
⚡ Reliability
25
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
72
Documentation
70
Error Messages
0
Auth Simplicity
60
Rate Limits
5

🔒 Security

TLS Enforcement
70
Auth Strength
65
Scope Granularity
20
Dep. Hygiene
45
Secret Handling
80

README suggests HTTPS/TLS is expected for deployments but does not explicitly state enforcement. Auth appears to be OAuth/OIDC-style via AUTH_* configuration, but there is no documented scope model/least-privilege and endpoint authorization requirements are not clear. Secrets are provided via environment variables (good practice), but there is no explicit statement about logging/redaction. Dependency hygiene cannot be confirmed from the provided content.

⚡ Reliability

Uptime/SLA
0
Version Stability
35
Breaking Changes
35
Error Recovery
30
AF Security Reliability

Best When

You want a self-hosted MCP stack in Go with an HTTP gateway and SSE-based event streaming, and you can provide/maintain the configuration and downstream LLM/service credentials.

Avoid When

You need a turnkey managed service with guaranteed uptime/support, or you require explicitly documented rate limits, error-code semantics, or idempotency/retry contracts at the API level.

Use Cases

  • Stand up an MCP server in Go to expose tool/resource capabilities over the Model Context Protocol
  • Build an MCP client that connects applications to an MCP server and streams events (SSE)
  • Create an HTTP-facing gateway that turns frontend requests into MCP client calls and optional LLM responses
  • Rapid prototyping/testing of MCP workflows using the included frontend project
  • Deploy basic observability for the MCP API/servers via Prometheus/Loki

Not For

  • Production deployments requiring well-specified operational guarantees (SLA, explicit retry/idempotency semantics) without additional review
  • Environments that require strict, documented data residency/compliance controls from the toolkit itself
  • Teams needing first-class, strongly documented SDKs beyond the Go ecosystem

Interface

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

Authentication

Methods: Environment-variable based auth configuration for the API/server (AUTH_DOMAIN, AUTH_CLIENT_ID, AUTH_CLIENT_SECRET, AUTH_CALLBACK_URL, AUTH_TOKEN_TTL, AUTH_AUDIENCE) Environment-variable based upstream credentials (e.g., GITHUB_TOKEN, ANTHROPIC_API_KEY)
OAuth: Yes Scopes: No

README indicates an auth setup via AUTH_* configuration (suggesting OAuth/OIDC-style flow), but does not document scopes/granularity or which endpoints require which auth. Upstream LLM/API keys are provided via environment variables.

Pricing

Free tier: No
Requires CC: No

Self-hosted open-source toolkit (MIT). Costs depend on infrastructure and any external LLM providers configured via environment variables.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • SSE/event-stream handling: agents must be prepared to consume streaming responses from the MCP server via the client/API layer
  • Auth requirements and endpoint-level permissions are not fully specified in the README; misconfigured AUTH_* variables may lead to opaque failures
  • Rate limits are not documented in the provided README; agents may need conservative request pacing
  • No documented idempotency/retry semantics are provided; replays may cause duplicate downstream calls depending on implementation

Alternatives

Full Evaluation Report

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

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 2026-04-04.

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