Iris

MCP-native agent evaluation and observability server — log traces, evaluate output quality, and track agent costs with 12 built-in eval rules and a real-time dashboard.

GitHub

Iris — The Agent Eval Standard for MCP

GitHub stars npm version npm downloads CI License: MIT

Know whether your AI agents are actually good enough to ship. Iris is an open-source MCP server that scores output quality, catches safety failures, and enforces cost budgets across all your agents. Any MCP-compatible agent discovers and uses it automatically — no SDK, no code changes.

Iris Dashboard

The Problem

Your agents are running in production. Infrastructure monitoring sees 200 OK and moves on. It has no idea the agent just:

  • Leaked a social security number in its response
  • Hallucinated an answer with zero factual grounding
  • Burned $0.47 on a single query — 4.7x your budget threshold
  • Made 6 tool calls when 2 would have sufficed

Iris evaluates all of it.

What You Get

Trace LoggingHierarchical span trees with per-tool-call latency, token usage, and cost in USD. Stored in SQLite, queryable instantly.
Output Evaluation12 built-in rules across 4 categories: completeness, relevance, safety, cost. PII detection, prompt injection patterns, hallucination markers. Add custom rules with Zod schemas.
Cost VisibilityAggregate cost across all agents over any time window. Set budget thresholds. Get flagged when agents overspend.
Web DashboardReal-time dark-mode UI with trace visualization, eval results, and cost breakdowns.

Quickstart

Add Iris to your Claude Desktop (or Cursor, Claude Code, Windsurf) MCP config:

{
  "mcpServers": {
    "iris-eval": {
      "command": "npx",
      "args": ["@iris-eval/mcp-server"]
    }
  }
}

That's it. Your agent discovers Iris and starts logging traces automatically.

Want the dashboard?

npx @iris-eval/mcp-server --dashboard
# Open http://localhost:6920

Other Install Methods

# Global install
npm install -g @iris-eval/mcp-server
iris-mcp --dashboard

# Docker
docker run -p 3000:3000 -v iris-data:/data ghcr.io/iris-eval/mcp-server

MCP Tools

Iris registers three tools that any MCP-compatible agent can invoke:

  • log_trace — Log an agent execution with spans, tool calls, token usage, and cost
  • evaluate_output — Score output quality against completeness, relevance, safety, and cost rules
  • get_traces — Query stored traces with filtering, pagination, and time-range support

Full tool schemas and configuration: iris-eval.com

Cloud Tier (Coming Soon)

Self-hosted Iris runs on your machine with SQLite. As your team's eval needs grow, the cloud tier adds PostgreSQL, team dashboards, alerting on quality regressions, and managed infrastructure.

Join the waitlist to get early access.

Examples

Community

Configuration & Security

CLI Arguments

FlagDefaultDescription
--transportstdioTransport type: stdio or http
--port3000HTTP transport port
--db-path~/.iris/iris.dbSQLite database path
--config~/.iris/config.jsonConfig file path
--api-key—API key for HTTP authentication
--dashboardfalseEnable web dashboard
--dashboard-port6920Dashboard port

Environment Variables

VariableDescription
IRIS_TRANSPORTTransport type
IRIS_PORTHTTP port
IRIS_DB_PATHDatabase path
IRIS_LOG_LEVELLog level: debug, info, warn, error
IRIS_DASHBOARDEnable dashboard (true/false)
IRIS_API_KEYAPI key for HTTP authentication
IRIS_ALLOWED_ORIGINSComma-separated allowed CORS origins

Security

When using HTTP transport, Iris includes:

  • API key authentication with timing-safe comparison
  • CORS restricted to localhost by default
  • Rate limiting (100 req/min API, 20 req/min MCP)
  • Helmet security headers
  • Zod input validation on all routes
  • ReDoS-safe regex for custom eval rules
  • 1MB request body limits
# Production deployment
iris-mcp --transport http --port 3000 --api-key "$(openssl rand -hex 32)" --dashboard

If Iris is useful to you, consider starring the repo — it helps others find it.

Star on GitHub

MIT Licensed.

相關伺服器