agent-eval-mcp
Provides CI-friendly quality gates for agent evaluation, enabling integration with GitHub Actions workflows for automated testing and deployment.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@agent-eval-mcp run evaluation suite on my candidate agent outputs"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Enterprise AI Agent Evaluation & Deployment Platform
A dependency-light evaluation platform for RAG/wiki-quality AI agents. It scores agent outputs for faithfulness, retrieval relevance, hallucination risk, latency, and cost; produces CI-friendly quality gates; emits regression/canary reports; and exposes the workflow through a lightweight MCP-style stdio tool server.
What Is Included
JSONL evaluation case format for RAG/wiki workflows.
Deterministic checks for:
faithfulness to retrieved context and reference answer,
retrieval relevance against question and expected keywords,
hallucination risk from unsupported answer content,
latency and cost thresholds.
100+ synthetic case generator.
CI/CD-style suite-level and case-level quality gates.
Markdown and JSON evaluation reports.
Regression report comparing baseline and candidate runs.
Canary promotion policy with traffic ramp decisions.
OpenTelemetry-compatible JSONL traces/metrics.
MCP-style stdio server exposing evaluation tools.
Related MCP server: mcp-llm-eval
Quick Start
git clone https://github.com/ad-github1/ENTERPRISE-AI-AGENT-EVALUATION-PLATFORM.git
cd ENTERPRISE-AI-AGENT-EVALUATION-PLATFORM
PYTHONPATH=src python3 -m agent_eval_platform generate-cases --count 120 --out examples/wiki_eval_cases.jsonl
PYTHONPATH=src python3 -m agent_eval_platform evaluate \
--cases examples/wiki_eval_cases.jsonl \
--gate examples/quality_gate.json \
--variant candidate \
--json-out reports/eval_result.json \
--markdown-out reports/eval_report.md \
--traces-out reports/traces.jsonl
PYTHONPATH=src python3 -m unittest discover -s testsTesting
Run the test suite:
PYTHONPATH=src python3 -m unittest discover -s testsCurrent local result:
Ran 5 tests in 0.015s
OKAfter installation:
pip install -e .
agent-eval evaluate --cases examples/wiki_eval_cases.jsonl --gate examples/quality_gate.json
agent-eval-mcpCase Format
Each JSONL row contains one evaluated agent run:
{
"case_id": "case-0001",
"question": "What contribution is Ada Lovelace known for in mathematics?",
"reference_answer": "Ada Lovelace is known for Analytical Engine notes.",
"expected_keywords": ["Ada Lovelace", "Analytical Engine"],
"retrieved_docs": [
{"doc_id": "wiki-1", "title": "Ada Lovelace", "text": "...", "score": 0.94}
],
"agent_answer": "Ada Lovelace is known for Analytical Engine notes.",
"latency_ms": 240.5,
"cost_usd": 0.0031,
"tags": ["wiki", "rag"]
}CI Quality Gate
The evaluator exits non-zero when --fail-on-gate is used and thresholds fail:
PYTHONPATH=src python3 -m agent_eval_platform evaluate \
--cases examples/wiki_eval_cases.jsonl \
--gate examples/quality_gate.json \
--fail-on-gateSee .github/workflows/agent-eval.yml for a GitHub Actions example.
MCP-Style Tool Server
Run:
PYTHONPATH=src python3 -m agent_eval_platform.mcp_serverSupported JSON-RPC methods:
initializetools/listtools/callwith:run_evaluation_suitecompare_regressiondecide_canary
This is intentionally stdio and dependency-free. It follows the MCP tool shape closely enough for local agent integration demos without requiring the MCP Python SDK.
Canary Workflow
PYTHONPATH=src python3 -m agent_eval_platform canary \
--result reports/eval_result.json \
--config examples/canary_config.json \
--json-out reports/canary_decision.jsonThe decision is hold, increase_traffic, or promote based on suite quality and minimum case coverage.
Evaluation Results
Evaluation Run
PYTHONPATH=src python3 -m agent_eval_platform evaluate \
--cases examples/wiki_eval_cases.jsonl \
--gate examples/quality_gate.json \
--variant candidate \
--json-out reports/eval_result.json \
--markdown-out reports/eval_report.md \
--traces-out reports/traces.jsonlAggregate Metrics
Metric | Value |
Evaluation cases | 120 |
Pass rate | 82.5% |
Average faithfulness | 0.825 |
Average retrieval relevance | 0.838 |
Average hallucination risk | 0.153 |
p50 latency | 392.58 ms |
p95 latency | 663.40 ms |
p99 latency | 872.87 ms |
Average cost | $0.00393 |
Total cost | $0.47158 |
Canary Decision
PYTHONPATH=src python3 -m agent_eval_platform canary \
--result reports/eval_result.json \
--config examples/canary_config.json \
--json-out reports/canary_decision.jsonResult:
{
"action": "hold",
"next_traffic_percent": 10.0,
"reasons": [
"pass_rate 0.825 < 0.900"
]
}The canary policy correctly blocked promotion because the candidate run did not meet the configured 90% pass-rate threshold. This demonstrates how the platform can prevent low-quality agent versions from being promoted automatically.
Observability
The evaluation emits OpenTelemetry-style JSONL traces to:
reports/traces.jsonlEach case generates spans for faithfulness, retrieval relevance, hallucination risk, and final case-level pass/fail status, enabling debugging of failed agent responses.
Available Tools
3 toolscompare_regressionC
Compare baseline and candidate suite JSON files.
Inputs: baseline_path (required), candidate_path (required), out.
decide_canaryC
Apply canary promotion policy to a suite result.
Inputs: config_path (required), result_path (required).
run_evaluation_suiteC
Run RAG/wiki agent evaluation cases and emit JSON, Markdown, and telemetry outputs.
Inputs: cases_path (required), variant, json_out, gate_path, +2 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 3 tools
Each tool has a distinct purpose: running evaluations, comparing results, and applying deployment policy. There is no overlap.
All tools follow a consistent verb_noun pattern (run_evaluation_suite, compare_regression, decide_canary), making them predictable.
With 3 tools, the server covers the core evaluation workflow without being bloated. Slightly minimal but appropriate for a focused server.
The tools cover the main lifecycle: run suite, compare results, decide promotion. Minor gaps like suite management exist, but the core workflow is complete.
Maintenance
Related MCP Connectors
MCP server for building and testing AI agents with multi-model experimentation and insights.
MCP server providing access to the Scorecard API to evaluate and optimize LLM systems.
Pay-per-call AI evaluation MCP server. Score LLM outputs against benchmark rubrics via Workers AI.
MCP server for AI agents to plan, verify, and deploy Cloudflare-native apps.
Related MCP Servers
- AlicenseAqualityBmaintenanceMCP server that lets coding agents test AI agents. Create YAML test cases, snapshot golden baselines, check for regressions, and generate visual reports all from inside Claude Code or any MCP-compatible tool. Works with LangGraph, CrewAI, OpenAI, Claude, Mistral, and any HTTP API.10143 npm1,798 PyPI142Apache 2.0
- AlicenseAqualityBmaintenanceA local MCP server that packages LLM evaluation gates as reusable CI/CD primitives, enabling AI agents to run datasets against models, score responses, and enforce quality thresholds.1079 PyPIMIT
- AlicenseAqualityDmaintenanceAn MCP server that lets any AI agent evaluate RAG outputs -- faithfulness scoring, hallucination detection, and retrieval quality metrics -- with zero API keys, using MCP sampling.6MIT
- AlicenseNot gradedqualityDmaintenanceA local MCP server that provides tools for AI coding agents to execute, evaluate, and diff regression tests for LLM agents.MIT