Mcpbr

What is Mcpbr about?

Mcpbr provides a single‑command harness to benchmark MCP servers using established software‑engineering datasets (SWE‑bench, CyberGym, MCPToolBench++). It runs parallel evaluations—one with the MCP server and one baseline—so you can measure the exact impact of tool access on task success.

How to use Mcpbr?

1. Install the package (pip or npm).
2. Set required environment variables (e.g., ANTHROPIC_API_KEY).
3. Initialise a configuration file with mcpbr init or edit an existing YAML.
4. Run the benchmark: mcpbr run -c <config.yaml> [options].
5. Review results in the console, JSON file, markdown report, or JUnit XML for CI integration.

Key features of Mcpbr

• Multiple benchmarks: SWE‑bench, CyberGym, MCPToolBench++ with configurable difficulty levels and sample sizes.
• Apples‑to‑apples comparison: identical model, dataset, and environment for MCP and baseline runs.
• Docker isolation: pre‑built images ensure reproducible environments; fallback to on‑the‑fly builds.
• Regression detection: compare against saved baseline results, set thresholds, and receive Slack/Discord/email alerts.
• Rich outputs: JSON, markdown, per‑task logs, and JUnit XML for CI pipelines.
• Extensible configuration: custom MCP server commands, timeouts, prompts, and model selection.
• CLI convenience: mcpbr models, mcpbr benchmarks, mcpbr cleanup, and verbose logging options.

Use cases of Mcpbr

• MCP server development – validate that new tool integrations actually improve LLM performance.
• CI/CD pipelines – automatically fail builds when regression thresholds are exceeded.
• Research – generate reproducible benchmark data for papers comparing tool‑augmented agents.
• Cost‑performance analysis – see whether tool usage justifies additional API costs.
• Product demos – quickly showcase quantitative gains of a proprietary MCP implementation.

FAQ from the Mcpbr

• Do I need Docker? Yes, Docker provides isolated environments for each benchmark task.
• Can I benchmark without a baseline? Use -M to run only the MCP evaluation or -B for baseline only.
• What models are supported? Currently Anthropic models (opus, sonnet, haiku) are listed via mcpbr models.
• How do I handle timeouts? Adjust timeout_seconds in the YAML or use startup_timeout_ms / tool_timeout_ms for MCP‑specific limits.
• Is there a pre‑built image for every task? Most common tasks have images; missing ones are built on the fly.
• Can I get JUnit output for GitHub Actions? Yes, use --output-junit <file>.xml and publish with actions like EnricoMi/publish-unit-test-result-action.
• How are API keys provided? Export them as environment variables (e.g., ANTHROPIC_API_KEY).

mcpbr

# Install via pip
pip install mcpbr && mcpbr init && mcpbr run -c mcpbr.yaml -n 1 -v

# Or via npm
npm install -g mcpbr-cli && mcpbr init && mcpbr run -c mcpbr.yaml -n 1 -v

Benchmark your MCP server against real GitHub issues. One command, hard numbers.

---

Model Context Protocol Benchmark Runner

Stop guessing if your MCP server actually helps. Get hard numbers comparing tool-assisted vs. baseline agent performance on real GitHub issues.

What You Get

Real metrics showing whether your MCP server improves agent performance on SWE-bench tasks. No vibes, just data.

Why mcpbr?

MCP servers promise to make LLMs better at coding tasks. But how do you prove it?

mcpbr runs controlled experiments: same model, same tasks, same environment - the only variable is your MCP server. You get:

• Apples-to-apples comparison against a baseline agent
• Real GitHub issues from SWE-bench (not toy examples)
• Reproducible results via Docker containers with pinned dependencies

Supported Benchmarks

mcpbr supports multiple software engineering benchmarks through a flexible abstraction layer:

SWE-bench (Default)

Real GitHub issues requiring bug fixes and patches. The agent generates unified diffs evaluated by running pytest test suites.

• Dataset: SWE-bench/SWE-bench_Lite
• Task: Generate patches to fix bugs
• Evaluation: Test suite pass/fail
• Pre-built images: Available for most tasks

CyberGym

Security vulnerabilities requiring Proof-of-Concept (PoC) exploits. The agent generates exploits that trigger crashes in vulnerable code.

• Dataset: sunblaze-ucb/cybergym
• Task: Generate PoC exploits
• Evaluation: PoC crashes pre-patch, doesn't crash post-patch
• Difficulty levels: 0-3 (controls context given to agent)
• Learn more: CyberGym Project

MCPToolBench++

Large-scale MCP tool use evaluation across 45+ categories. Tests agent capabilities in tool discovery, selection, invocation, and result interpretation.

• Dataset: MCPToolBench/MCPToolBenchPP
• Task: Complete tasks using appropriate MCP tools
• Evaluation: Tool selection accuracy, parameter correctness, sequence matching
• Categories: Browser, Finance, Code Analysis, and 40+ more
• Learn more: MCPToolBench++ Paper | GitHub

# Run SWE-bench (default)
mcpbr run -c config.yaml

# Run CyberGym at level 2
mcpbr run -c config.yaml --benchmark cybergym --level 2

# Run MCPToolBench++
mcpbr run -c config.yaml --benchmark mcptoolbench

# List available benchmarks
mcpbr benchmarks

See the benchmarks guide for details on each benchmark and how to configure them.

Overview

This harness runs two parallel evaluations for each task:

1. MCP Agent: LLM with access to tools from your MCP server
2. Baseline Agent: LLM without tools (single-shot generation)

By comparing these, you can measure the effectiveness of your MCP server for different software engineering tasks. See the MCP integration guide for tips on testing your server.

Regression Detection

mcpbr includes built-in regression detection to catch performance degradations between MCP server versions:

Key Features

• Automatic Detection: Compare current results against a baseline to identify regressions
• Detailed Reports: See exactly which tasks regressed and which improved
• Threshold-Based Exit Codes: Fail CI/CD pipelines when regression rate exceeds acceptable limits
• Multi-Channel Alerts: Send notifications via Slack, Discord, or email

How It Works

A regression is detected when a task that passed in the baseline now fails in the current run. This helps you catch issues before deploying new versions of your MCP server.

# First, run a baseline evaluation and save results
mcpbr run -c config.yaml -o baseline.json

# Later, compare a new version against the baseline
mcpbr run -c config.yaml --baseline-results baseline.json --regression-threshold 0.1

# With notifications
mcpbr run -c config.yaml --baseline-results baseline.json \
  --regression-threshold 0.1 \
  --slack-webhook https://hooks.slack.com/services/YOUR/WEBHOOK/URL

Use Cases

• CI/CD Integration: Automatically detect regressions in pull requests
• Version Comparison: Compare different versions of your MCP server
• Performance Monitoring: Track MCP server performance over time
• Team Notifications: Alert your team when regressions are detected

Example Output

======================================================================
REGRESSION DETECTION REPORT
======================================================================

Total tasks compared: 25
Regressions detected: 2
Improvements detected: 5
Regression rate: 8.0%

REGRESSIONS (previously passed, now failed):
----------------------------------------------------------------------
  - django__django-11099
    Error: Timeout
  - sympy__sympy-18087
    Error: Test suite failed

IMPROVEMENTS (previously failed, now passed):
----------------------------------------------------------------------
  - astropy__astropy-12907
  - pytest-dev__pytest-7373
  - scikit-learn__scikit-learn-25570
  - matplotlib__matplotlib-23913
  - requests__requests-3362

======================================================================

For CI/CD integration, use --regression-threshold to fail the build when regressions exceed an acceptable rate:

# .github/workflows/test-mcp.yml
- name: Run mcpbr with regression detection
  run: |
    mcpbr run -c config.yaml \
      --baseline-results baseline.json \
      --regression-threshold 0.1 \
      -o current.json

This will exit with code 1 if the regression rate exceeds 10%, failing the CI job.

Installation

Full installation guide with detailed setup instructions.

• Python 3.11+
• Docker (running)
• ANTHROPIC_API_KEY environment variable
• Claude Code CLI (claude) installed
• Network access (for pulling Docker images and API calls)

Supported Models (aliases or full names):

• Claude Opus 4.5: opus or claude-opus-4-5-20251101
• Claude Sonnet 4.5: sonnet or claude-sonnet-4-5-20250929
• Claude Haiku 4.5: haiku or claude-haiku-4-5-20251001

Run mcpbr models to see the full list.

via npm

# Run with npx (no installation)
npx mcpbr-cli run -c config.yaml

# Or install globally
npm install -g mcpbr-cli
mcpbr run -c config.yaml

Note: The npm package requires Python 3.11+ and the mcpbr Python package (pip install mcpbr)

via pip

# Install from PyPI
pip install mcpbr

# Or install from source
git clone https://github.com/greynewell/mcpbr.git
cd mcpbr
pip install -e .

# Or with uv
uv pip install -e .

Note for Apple Silicon users: The harness automatically uses x86_64 Docker images via emulation. This may be slower than native ARM64 images but ensures compatibility with all SWE-bench tasks.

Quick Start

Option 1: Use Example Configurations (Recommended)

Get started in seconds with our example configurations:

# Set your API key
export ANTHROPIC_API_KEY="your-api-key"

# Run your first evaluation using an example config
mcpbr run -c examples/quick-start/getting-started.yaml -v

This runs 5 SWE-bench tasks with the filesystem server. Expected runtime: 15-30 minutes, cost: $2-5.

Explore 25+ example configurations in the examples/ directory:

• Quick Start: Getting started, testing servers, comparing models
• Benchmarks: SWE-bench Lite/Full, CyberGym basic/advanced
• MCP Servers: Filesystem, GitHub, Brave Search, databases, custom servers
• Scenarios: Cost-optimized, performance-optimized, CI/CD, regression detection

See the Examples README for the complete guide.

Option 2: Generate Custom Configuration

1. Set your API key:

export ANTHROPIC_API_KEY="your-api-key"

2. Generate a configuration file:

mcpbr init

3. Edit the configuration to point to your MCP server:

mcp_server:
  command: "npx"
  args:
    - "-y"
    - "@modelcontextprotocol/server-filesystem"
    - "{workdir}"
  env: {}

provider: "anthropic"
agent_harness: "claude-code"

model: "sonnet"  # or full name: "claude-sonnet-4-5-20250929"
dataset: "SWE-bench/SWE-bench_Lite"
sample_size: 10
timeout_seconds: 300
max_concurrent: 4

4. Run the evaluation:

mcpbr run --config config.yaml

Claude Code Integration

mcpbr includes a built-in Claude Code plugin that makes Claude an expert at running benchmarks correctly. When you clone this repository, Claude Code automatically detects the plugin and gains specialized knowledge about mcpbr.

What This Means for You

When using Claude Code in this repository, you can simply say:

• "Run the SWE-bench Lite benchmark"
• "Generate a config for my MCP server"
• "Run a quick test with 1 task"

Claude will automatically:

• Verify Docker is running before starting
• Check for required API keys
• Generate valid configurations with proper {workdir} placeholders
• Use correct CLI flags and options
• Provide helpful troubleshooting when issues occur

Available Skills

The plugin includes three specialized skills:

1. run-benchmark: Expert at running evaluations with proper validation

   • Checks prerequisites (Docker, API keys, config files)
   • Constructs valid mcpbr run commands
   • Handles errors gracefully with actionable feedback

2. generate-config: Generates valid mcpbr configuration files

   • Ensures {workdir} placeholder is included
   • Validates MCP server commands
   • Provides benchmark-specific templates

3. swe-bench-lite: Quick-start command for SWE-bench Lite

   • Pre-configured for 5-task evaluation
   • Includes sensible defaults for output files
   • Perfect for testing and demonstrations

Getting Started with Claude Code

Just clone the repository and start asking Claude to run benchmarks:

git clone https://github.com/greynewell/mcpbr.git
cd mcpbr

# In Claude Code, simply say:
# "Run the SWE-bench Lite eval with 5 tasks"

The bundled plugin ensures Claude makes no silly mistakes and follows best practices automatically.

Configuration

Full configuration reference with all options and examples.

MCP Server Configuration

The mcp_server section defines how to start your MCP server:

Field	Description
command	Executable to run (e.g., npx, uvx, python)
args	Command arguments. Use {workdir} as placeholder for the task repository path
env	Additional environment variables

Example Configurations

Anthropic Filesystem Server:

mcp_server:
  command: "npx"
  args: ["-y", "@modelcontextprotocol/server-filesystem", "{workdir}"]

Custom Python MCP Server:

mcp_server:
  command: "python"
  args: ["-m", "my_mcp_server", "--workspace", "{workdir}"]
  env:
    LOG_LEVEL: "debug"

Supermodel Codebase Analysis Server:

mcp_server:
  command: "npx"
  args: ["-y", "@supermodeltools/mcp-server"]
  env:
    SUPERMODEL_API_KEY: "${SUPERMODEL_API_KEY}"

MCP Timeout Configuration

mcpbr supports configurable timeouts for MCP server operations to handle different server types and workloads.

Configuration Fields

Field	Description	Default
startup_timeout_ms	Timeout in milliseconds for MCP server startup	60000 (60s)
tool_timeout_ms	Timeout in milliseconds for MCP tool execution	900000 (15 min)

These fields map to the MCP_TIMEOUT and MCP_TOOL_TIMEOUT environment variables used by Claude Code. See the Claude Code settings documentation for more details.

Example Configuration

mcp_server:
  command: "npx"
  args: ["-y", "@modelcontextprotocol/server-filesystem", "{workdir}"]
  startup_timeout_ms: 60000      # 60 seconds for server to start
  tool_timeout_ms: 900000        # 15 minutes for long-running tools

Common Timeout Values

Different server types require different timeout settings based on their operational characteristics:

Server Type	startup_timeout_ms	tool_timeout_ms	Notes
Fast (filesystem, git)	10000 (10s)	30000 (30s)	Local operations with minimal overhead
Medium (web search, APIs)	30000 (30s)	120000 (2m)	Network I/O with moderate latency
Slow (code analysis, databases)	60000 (60s)	900000 (15m)	Complex processing or large datasets

When to adjust timeouts:

• Increase startup_timeout_ms if your server takes longer to initialize (e.g., loading large models, establishing database connections)
• Increase tool_timeout_ms if your tools perform long-running operations (e.g., codebase analysis, file processing, AI inference)
• Decrease timeouts for fast servers to fail quickly on connection issues

Custom Agent Prompt

You can customize the prompt sent to the agent using the agent_prompt field:

agent_prompt: |
  Fix the following bug in this repository:

  {problem_statement}

  Make the minimal changes necessary to fix the issue.
  Focus on the root cause, not symptoms.

Use {problem_statement} as a placeholder for the SWE-bench issue text. You can also override the prompt via CLI with --prompt.

Evaluation Parameters

Parameter	Default	Description
provider	anthropic	LLM provider
agent_harness	claude-code	Agent backend
benchmark	swe-bench	Benchmark to run (swe-bench, cybergym, or mcptoolbench)
agent_prompt	null	Custom prompt template (use {problem_statement} placeholder)
model	sonnet	Model alias or full ID
dataset	null	HuggingFace dataset (optional, benchmark provides default)
cybergym_level	1	CyberGym difficulty level (0-3, only for CyberGym benchmark)
sample_size	null	Number of tasks (null = full dataset)
timeout_seconds	300	Timeout per task
max_concurrent	4	Parallel task limit
max_iterations	10	Max agent iterations per task

CLI Reference

Full CLI documentation with all commands and options.

Get help for any command with --help or -h:

mcpbr --help
mcpbr run --help
mcpbr init --help

Commands Overview

Command	Description
mcpbr run	Run benchmark evaluation with configured MCP server
mcpbr init	Generate an example configuration file
mcpbr models	List supported models for evaluation
mcpbr providers	List available model providers
mcpbr harnesses	List available agent harnesses
mcpbr benchmarks	List available benchmarks (SWE-bench, CyberGym, MCPToolBench++)
mcpbr cleanup	Remove orphaned mcpbr Docker containers

mcpbr run

Run SWE-bench evaluation with the configured MCP server.

Option	Short	Description
--config PATH	-c	Path to YAML configuration file (required)
--model TEXT	-m	Override model from config
--benchmark TEXT	-b	Override benchmark from config (swe-bench, cybergym, or mcptoolbench)
--level INTEGER		Override CyberGym difficulty level (0-3)
--sample INTEGER	-n	Override sample size from config
--mcp-only	-M	Run only MCP evaluation (skip baseline)
--baseline-only	-B	Run only baseline evaluation (skip MCP)
--no-prebuilt		Disable pre-built SWE-bench images (build from scratch)
--output PATH	-o	Path to save JSON results
--report PATH	-r	Path to save Markdown report
--output-junit PATH		Path to save JUnit XML report (for CI/CD integration)
--verbose	-v	Verbose output (-v summary, -vv detailed)
--log-file PATH	-l	Path to write raw JSON log output (single file)
--log-dir PATH		Directory to write per-instance JSON log files
--task TEXT	-t	Run specific task(s) by instance_id (repeatable)
--prompt TEXT		Override agent prompt (use {problem_statement} placeholder)
--baseline-results PATH		Path to baseline results JSON for regression detection
--regression-threshold FLOAT		Maximum acceptable regression rate (0-1). Exit with code 1 if exceeded.
--slack-webhook URL		Slack webhook URL for regression notifications
--discord-webhook URL		Discord webhook URL for regression notifications
--email-to EMAIL		Email address for regression notifications
--email-from EMAIL		Sender email address for notifications
--smtp-host HOST		SMTP server hostname for email notifications
--smtp-port PORT		SMTP server port (default: 587)
--smtp-user USER		SMTP username for authentication
--smtp-password PASS		SMTP password for authentication
--help	-h	Show help message

# Full evaluation (MCP + baseline)
mcpbr run -c config.yaml

# Run only MCP evaluation
mcpbr run -c config.yaml -M

# Run only baseline evaluation
mcpbr run -c config.yaml -B

# Override model
mcpbr run -c config.yaml -m claude-3-5-sonnet-20241022

# Override sample size
mcpbr run -c config.yaml -n 50

# Save results and report
mcpbr run -c config.yaml -o results.json -r report.md

# Save JUnit XML for CI/CD
mcpbr run -c config.yaml --output-junit junit.xml

# Run specific tasks
mcpbr run -c config.yaml -t astropy__astropy-12907 -t django__django-11099

# Verbose output with per-instance logs
mcpbr run -c config.yaml -v --log-dir logs/

# Very verbose output
mcpbr run -c config.yaml -vv

# Run CyberGym benchmark
mcpbr run -c config.yaml --benchmark cybergym --level 2

# Run CyberGym with specific tasks
mcpbr run -c config.yaml --benchmark cybergym --level 3 -n 5

# Regression detection - compare against baseline
mcpbr run -c config.yaml --baseline-results baseline.json

# Regression detection with threshold (exit 1 if exceeded)
mcpbr run -c config.yaml --baseline-results baseline.json --regression-threshold 0.1

# Regression detection with Slack notifications
mcpbr run -c config.yaml --baseline-results baseline.json --slack-webhook https://hooks.slack.com/...

# Regression detection with Discord notifications
mcpbr run -c config.yaml --baseline-results baseline.json --discord-webhook https://discord.com/api/webhooks/...

# Regression detection with email notifications
mcpbr run -c config.yaml --baseline-results baseline.json \
  --email-to team@example.com --email-from mcpbr@example.com \
  --smtp-host smtp.gmail.com --smtp-port 587 \
  --smtp-user user@gmail.com --smtp-password "app-password"

mcpbr init

Generate an example configuration file.

Option	Short	Description
--output PATH	-o	Path to write example config (default: mcpbr.yaml)
--help	-h	Show help message

mcpbr init
mcpbr init -o my-config.yaml

mcpbr models

List supported Anthropic models for evaluation.

mcpbr cleanup

Remove orphaned mcpbr Docker containers that were not properly cleaned up.

Option	Short	Description
--dry-run		Show containers that would be removed without removing them
--force	-f	Skip confirmation prompt
--help	-h	Show help message

# Preview containers to remove
mcpbr cleanup --dry-run

# Remove containers with confirmation
mcpbr cleanup

# Remove containers without confirmation
mcpbr cleanup -f

Example Run

Here's what a typical evaluation looks like:

$ mcpbr run -c config.yaml -v -o results.json --log-dir my-logs

mcpbr Evaluation
  Config: config.yaml
  Provider: anthropic
  Model: sonnet
  Agent Harness: claude-code
  Dataset: SWE-bench/SWE-bench_Lite
  Sample size: 10
  Run MCP: True, Run Baseline: True
  Pre-built images: True
  Log dir: my-logs

Loading dataset: SWE-bench/SWE-bench_Lite
Evaluating 10 tasks
Provider: anthropic, Harness: claude-code
14:23:15 [MCP] Starting mcp run for astropy-12907:mcp
14:23:22 astropy-12907:mcp    > TodoWrite
14:23:22 astropy-12907:mcp    < Todos have been modified successfully...
14:23:26 astropy-12907:mcp    > Glob
14:23:26 astropy-12907:mcp    > Grep
14:23:27 astropy-12907:mcp    < $WORKDIR/astropy/modeling/separable.py
14:23:27 astropy-12907:mcp    < Found 5 files: astropy/modeling/tests/test_separable.py...
...
14:27:43 astropy-12907:mcp    * done turns=31 tokens=115/6,542
14:28:30 [BASELINE] Starting baseline run for astropy-12907:baseline
...

Output

Understanding evaluation results - detailed guide to interpreting output.

Console Output

The harness displays real-time progress with verbose mode (-v) and a final summary table:

Evaluation Results

                 Summary
+-----------------+-----------+----------+
| Metric          | MCP Agent | Baseline |
+-----------------+-----------+----------+
| Resolved        | 8/25      | 5/25     |
| Resolution Rate | 32.0%     | 20.0%    |
+-----------------+-----------+----------+

Improvement: +60.0%

Per-Task Results
+------------------------+------+----------+-------+
| Instance ID            | MCP  | Baseline | Error |
+------------------------+------+----------+-------+
| astropy__astropy-12907 | PASS |   PASS   |       |
| django__django-11099   | PASS |   FAIL   |       |
| sympy__sympy-18087     | FAIL |   FAIL   |       |
+------------------------+------+----------+-------+

Results saved to results.json

JSON Output (--output)

{
  "metadata": {
    "timestamp": "2026-01-17T07:23:39.871437+00:00",
    "config": {
      "model": "sonnet",
      "provider": "anthropic",
      "agent_harness": "claude-code",
      "dataset": "SWE-bench/SWE-bench_Lite",
      "sample_size": 25,
      "timeout_seconds": 600,
      "max_iterations": 30
    },
    "mcp_server": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "{workdir}"]
    }
  },
  "summary": {
    "mcp": {"resolved": 8, "total": 25, "rate": 0.32},
    "baseline": {"resolved": 5, "total": 25, "rate": 0.20},
    "improvement": "+60.0%"
  },
  "tasks": [
    {
      "instance_id": "astropy__astropy-12907",
      "mcp": {
        "patch_generated": true,
        "tokens": {"input": 115, "output": 6542},
        "iterations": 30,
        "tool_calls": 72,
        "tool_usage": {
          "TodoWrite": 4, "Task": 1, "Glob": 4,
          "Grep": 11, "Bash": 27, "Read": 22,
          "Write": 2, "Edit": 1
        },
        "resolved": true,
        "patch_applied": true,
        "fail_to_pass": {"passed": 2, "total": 2},
        "pass_to_pass": {"passed": 10, "total": 10}
      },
      "baseline": {
        "patch_generated": true,
        "tokens": {"input": 63, "output": 7615},
        "iterations": 30,
        "tool_calls": 57,
        "tool_usage": {
          "TodoWrite": 4, "Glob": 3, "Grep": 4,
          "Read": 14, "Bash": 26, "Write": 4, "Edit": 1
        },
        "resolved": true,
        "patch_applied": true
      }
    }
  ]
}

Markdown Report (--report)

Generates a human-readable report with:

• Summary statistics
• Per-task results table
• Analysis of which tasks each agent solved

Per-Instance Logs (--log-dir)

Creates a directory with detailed JSON log files for each task run. Filenames include timestamps to prevent overwrites:

my-logs/
  astropy__astropy-12907_mcp_20260117_143052.json
  astropy__astropy-12907_baseline_20260117_143156.json
  django__django-11099_mcp_20260117_144023.json
  django__django-11099_baseline_20260117_144512.json

Each log file contains the full stream of events from the agent CLI:

{
  "instance_id": "astropy__astropy-12907",
  "run_type": "mcp",
  "events": [
    {
      "type": "system",
      "subtype": "init",
      "cwd": "/workspace",
      "tools": ["Task", "Bash", "Glob", "Grep", "Read", "Edit", "Write", "TodoWrite"],
      "model": "claude-sonnet-4-5-20250929",
      "claude_code_version": "2.1.12"
    },
    {
      "type": "assistant",
      "message": {
        "content": [{"type": "text", "text": "I'll help you fix this bug..."}]
      }
    },
    {
      "type": "assistant",
      "message": {
        "content": [{"type": "tool_use", "name": "Grep", "input": {"pattern": "separability"}}]
      }
    },
    {
      "type": "result",
      "num_turns": 31,
      "usage": {"input_tokens": 115, "output_tokens": 6542}
    }
  ]
}

This is useful for debugging failed runs or analyzing agent behavior in detail.

JUnit XML Output (--output-junit)

The harness can generate JUnit XML reports for integration with CI/CD systems like GitHub Actions, GitLab CI, and Jenkins. Each task is represented as a test case, with resolved/unresolved tasks mapped to pass/fail states.

mcpbr run -c config.yaml --output-junit junit.xml

The JUnit XML report includes:

• Test Suites: Separate suites for MCP and baseline evaluations
• Test Cases: Each task is a test case with timing information
• Failures: Unresolved tasks with detailed error messages
• Properties: Metadata about model, provider, benchmark configuration
• System Output: Token usage, tool calls, and test results per task

CI/CD Integration Examples

GitHub Actions:

name: MCP Benchmark

on: [push, pull_request]

jobs:
  benchmark:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3

      - name: Set up Python
        uses: actions/setup-python@v4
        with:
          python-version: '3.11'

      - name: Install mcpbr
        run: pip install mcpbr

      - name: Run benchmark
        env:
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
        run: |
          mcpbr run -c config.yaml --output-junit junit.xml

      - name: Publish Test Results
        uses: EnricoMi/publish-unit-test-result-action@v2
        if: always()
        with:
          files: junit.xml

GitLab CI:

benchmark:
  image: python:3.11
  services:
    - docker:dind
  script:
    - pip install mcpbr
    - mcpbr run -c config.yaml --output-junit junit.xml
  artifacts:
    reports:
      junit: junit.xml

Jenkins:

pipeline {
    agent any
    stages {
        stage('Benchmark') {
            steps {
                sh 'pip install mcpbr'
                sh 'mcpbr run -c config.yaml --output-junit junit.xml'
            }
        }
    }
    post {
        always {
            junit 'junit.xml'
        }
    }
}

The JUnit XML format enables native test result visualization in your CI/CD dashboard, making it easy to track benchmark performance over time and identify regressions.

How It Works

Architecture deep dive - learn how mcpbr works internally.

1. Load Tasks: Fetches tasks from the selected benchmark (SWE-bench, CyberGym, or MCPToolBench++) via HuggingFace
2. Create Environment: For each task, creates an isolated Docker environment with the repository and dependencies
3. Run MCP Agent: Invokes Claude Code CLI inside the Docker container, letting it explore and generate a solution (patch or PoC)
4. Run Baseline: Same as MCP agent but without the MCP server
5. Evaluate: Runs benchmark-specific evaluation (test suites for SWE-bench, crash detection for CyberGym, tool use accuracy for MCPToolBench++)
6. Report: Aggregates results and calculates improvement

Pre-built Docker Images

The harness uses pre-built SWE-bench Docker images from Epoch AI's registry when available. These images come with:

• The repository checked out at the correct commit
• All project dependencies pre-installed and validated
• A consistent environment for reproducible evaluations

The agent (Claude Code CLI) runs inside the container, which means:

• Python imports work correctly (e.g., from astropy import ...)
• The agent can run tests and verify fixes
• No dependency conflicts with the host machine

If a pre-built image is not available for a task, the harness falls back to cloning the repository and attempting to install dependencies (less reliable).

Architecture

mcpbr/
├── src/mcpbr/
│   ├── cli.py           # Command-line interface
│   ├── config.py        # Configuration models
│   ├── models.py        # Supported model registry
│   ├── providers.py     # LLM provider abstractions (extensible)
│   ├── harnesses.py     # Agent harness implementations (extensible)
│   ├── benchmarks/      # Benchmark abstraction layer
│   │   ├── __init__.py      # Registry and factory
│   │   ├── base.py          # Benchmark protocol
│   │   ├── swebench.py      # SWE-bench implementation
│   │   ├── cybergym.py      # CyberGym implementation
│   │   └── mcptoolbench.py  # MCPToolBench++ implementation
│   ├── harness.py       # Main orchestrator
│   ├── agent.py         # Baseline agent implementation
│   ├── docker_env.py    # Docker environment management + in-container execution
│   ├── evaluation.py    # Patch application and testing
│   ├── log_formatter.py # Log formatting and per-instance logging
│   └── reporting.py     # Output formatting
├── tests/
│   ├── test_*.py        # Unit tests
│   ├── test_benchmarks.py # Benchmark tests
│   └── test_integration.py  # Integration tests
├── Dockerfile           # Fallback image for task environments
└── config/
    └── example.yaml     # Example configuration

The architecture uses Protocol-based abstractions for providers, harnesses, and benchmarks, making it easy to add support for additional LLM providers, agent backends, or software engineering benchmarks in the future. See the API reference and benchmarks guide for more details.

Execution Flow

┌─────────────────────────────────────────────────────────────────┐
│                         Host Machine                            │
│  ┌───────────────────────────────────────────────────────────┐  │
│  │                    mcpbr Harness (Python)                 │  │
│  │  - Loads SWE-bench tasks from HuggingFace                 │  │
│  │  - Pulls pre-built Docker images                          │  │
│  │  - Orchestrates agent runs                                │  │
│  │  - Collects results and generates reports                 │  │
│  └─────────────────────────┬─────────────────────────────────┘  │
│                            │ docker exec                        │
│  ┌─────────────────────────▼─────────────────────────────────┐  │
│  │              Docker Container (per task)                  │  │
│  │  ┌─────────────────────────────────────────────────────┐  │  │
│  │  │  Pre-built SWE-bench Image                          │  │  │
│  │  │  - Repository at correct commit                     │  │  │
│  │  │  - All dependencies installed (astropy, django...)  │  │  │
│  │  │  - Node.js + Claude CLI (installed at startup)      │  │  │
│  │  └─────────────────────────────────────────────────────┘  │  │
│  │                                                           │  │
│  │  Agent (Claude Code CLI) runs HERE:                       │  │
│  │  - Makes API calls to Anthropic                           │  │
│  │  - Executes Bash commands (with working imports!)         │  │
│  │  - Reads/writes files                                     │  │
│  │  - Generates patches                                      │  │
│  │                                                           │  │
│  │  Evaluation runs HERE:                                    │  │
│  │  - Applies patch via git                                  │  │
│  │  - Runs pytest with task's test suite                     │  │
│  └───────────────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────────────┘

Troubleshooting

FAQ - Quick answers to common questions

Full troubleshooting guide - Detailed solutions to common issues

Docker Issues

Ensure Docker is running:

docker info

Pre-built Image Not Found

If the harness can't pull a pre-built image for a task, it will fall back to building from scratch. You can also manually pull images:

docker pull ghcr.io/epoch-research/swe-bench.eval.x86_64.astropy__astropy-12907

Slow on Apple Silicon

On ARM64 Macs, x86_64 Docker images run via emulation which is slower. This is normal. If you're experiencing issues, ensure you have Rosetta 2 installed:

softwareupdate --install-rosetta

MCP Server Not Starting

Test your MCP server independently:

npx -y @modelcontextprotocol/server-filesystem /tmp/test

API Key Issues

Ensure your Anthropic API key is set:

export ANTHROPIC_API_KEY="sk-ant-..."

Timeout Issues

Increase the timeout in your config:

timeout_seconds: 600

Claude CLI Not Found

Ensure the Claude Code CLI is installed and in your PATH:

which claude  # Should return the path to the CLI

Development

# Install dev dependencies
pip install -e ".[dev]"

# Run unit tests
pytest -m "not integration"

# Run integration tests (requires API keys and Docker)
pytest -m integration

# Run all tests
pytest

# Lint
ruff check src/

Roadmap

We're building the defacto standard for MCP server benchmarking! Our v1.0 Roadmap includes 200+ features across 11 strategic categories:

🎯 Good First Issues | 🙋 Help Wanted | 📋 View Roadmap

Roadmap Highlights

Phase 1: Foundation (v0.3.0)

• ✅ JUnit XML output format for CI/CD integration
• CSV, YAML, XML output formats
• Config validation and templates
• Results persistence and recovery
• Cost analysis in reports

Phase 2: Benchmarks (v0.4.0)

• HumanEval, MBPP, ToolBench
• GAIA for general AI capabilities
• Custom benchmark YAML support
• SWE-bench Verified

Phase 3: Developer Experience (v0.5.0)

• Real-time dashboard
• Interactive config wizard
• Shell completion
• Pre-flight checks

Phase 4: Platform Expansion (v0.6.0)

• NPM package
• GitHub Action for CI/CD
• Homebrew formula
• Official Docker image

Phase 5: MCP Testing Suite (v1.0.0)

• Tool coverage analysis
• Performance profiling
• Error rate monitoring
• Security scanning

Get Involved

We welcome contributions! Check out our 30+ good first issues perfect for newcomers:

• Output Formats: CSV/YAML/XML export
• Configuration: Validation, templates, shell completion
• Platform: Homebrew formula, Conda package
• Documentation: Best practices, examples, guides

See the contributing guide to get started!

Best Practices

New to mcpbr or want to optimize your workflow? Check out the Best Practices Guide for:

• Benchmark selection guidelines
• MCP server configuration tips
• Performance optimization strategies
• Cost management techniques
• CI/CD integration patterns
• Debugging workflows
• Common pitfalls to avoid

Contributing

Please see CONTRIBUTING.md or the contributing guide for guidelines on how to contribute.

All contributors are expected to follow our Community Guidelines.

License

MIT - see LICENSE for details.

---

Built by Grey Newell