iris-eval

v0.18.0

Stop shipping agents on vibes. Score every agent output for quality, safety, and cost.

Claude Code1 Skill1 MCP serverstdio

By IrisLicense: MIT9 GitHub starsUpdated 1 hour ago

Directory evidence

Runtimes: Claude Code
Parsed components: 2 skill or MCP entries
Source updated: Sep 24, 2026
Manifest status: Canonical path parsed

The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →

Install iris-eval for Claude Code

Installs for the current user

claude plugin marketplace add IchenDEV/agent-plugin-mkt
claude plugin marketplace update agent-plugin-marketplace
claude plugin install iris-eval@agent-plugin-marketplace

Paste and run these commands in a terminal with Claude Code. They add and refresh the PluginsMP catalog, then install this plugin.

The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.

Get the source manually

$ git clone https://github.com/iris-eval/mcp-server

Clone the source repository, then follow its setup instructions to add the plugin to a compatible client. The repository root is the plugin root.

Plugin files

iris-eval/

├── .claude-plugin/plugin.json
├── skills/iris-eval/SKILL.md
└── .mcp.json

Included Skills1

iris-evalskills/iris-eval/SKILL.md

Evaluate AI agent outputs for quality, safety, and cost using the Iris MCP server. Use when reviewing agent responses, checking for PII leaks, scoring output quality, or tracking execution costs.

MCP servers1

iris-evalstdio

command: npx
args: -y @iris-eval/mcp-server
env.IRIS_LOG_LEVEL: info

Plugin manifests1

.claude-plugin/plugin.json

{
  "name": "iris-eval",
  "version": "0.18.0",
  "description": "Stop shipping agents on vibes. Score every agent output for quality, safety, and cost.",
  "author": {
    "name": "Iris",
    "email": "[email protected]",
    "url": "https://github.com/iris-eval"
  },
  "homepage": "https://iris-eval.com",
  "repository": "https://github.com/iris-eval/mcp-server",
  "license": "MIT",
  "keywords": [
    "eval",
    "evaluation",
    "mcp",
    "quality",
    "safety",
    "cost",
    "agent",
    "scoring"
  ]
}

Browse the “eval” tag →

iris-eval

by Iris

Stop shipping agents on vibes — agent eval for any MCP workflow. Score output quality, catch PII and prompt injection, verify citations, and enforce cost budgets. Adds the Iris MCP server (12 tools) plus the iris-eval skill that guides eval-driven development.

Claude Code1 Skill1 MCP serverstdio

9Updated 1 hour ago

iris-eval-capture

by Iris

Record every Claude Code turn into Iris without the model calling a tool: the prompt, the tool calls and the final answer, stored and evaluated locally, with critical spans redacted in the stored evaluation text. Opt-in; install beside the iris-eval plugin or on its own. Needs @iris-eval/mcp-server 0.18.0 on the machine (npx fetches it once).

Claude Code

9Updated 1 hour ago

ai-rag-engineering

by Matt Corbett

AI / RAG Engineering specialist team — agents (rag-architect-lead, retrieval-eval-analyst, ingestion-chunking-specialist, llm-serving-cost-specialist), skills, templates, commands, an advisory hook, best-practice rules, and a research-grounded knowledge bank. A RAG team for an ML engineer or AI product lead accountable for answer quality and serving cost — it fixes retrieval before generation, treats chunking as a retrieval decision, evals before it ships, and reads context-window and token economics. Inherits ravenclaude-core protocols.

Claude Code5 Skills

7Updated last week

For maintainers

If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.

[iris-eval on Agent Plugins Marketplace](https://pluginsmp.com/plugins/iris-eval)