Code Runner MCP Server

Code Runner MCP server enabling AI agents to execute code in multiple programming languages — running Python, JavaScript, TypeScript, Bash, Go, Rust, and other language snippets, capturing stdout/stderr output, and integrating local code execution into agent-driven development, testing, and data analysis workflows.

Evaluated Mar 06, 2026 (0d ago) vcurrent
Homepage ↗ Repo ↗ Developer Tools code-execution code-runner mcp-server sandbox polyglot testing
⚙ Agent Friendliness
79
/ 100
Can an agent use this?
🔒 Security
72
/ 100
Is it safe for agents?
⚡ Reliability
70
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
72
Documentation
75
Error Messages
72
Auth Simplicity
98
Rate Limits
90

🔒 Security

TLS Enforcement
78
Auth Strength
65
Scope Granularity
62
Dep. Hygiene
70
Secret Handling
85

CRITICAL: Full host access — no sandboxing. Only trusted agents. No credentials needed. Trusted community author (formulahendry). Use containerized alternative for untrusted code.

⚡ Reliability

Uptime/SLA
72
Version Stability
70
Breaking Changes
70
Error Recovery
70
AF Security Reliability

Best When

An agent needs to run code locally to verify, test, or process data — for development workflows where the agent can generate and immediately validate code on the local machine.

Avoid When

You need isolated/sandboxed execution for untrusted code — local execution has full system access. Use containerized alternatives for security.

Use Cases

  • Executing Python scripts for data processing from analytics agents
  • Running code snippets to verify solutions from coding assistant agents
  • Testing algorithms and logic from software development agents
  • Executing shell commands and scripts from automation agents
  • Running unit tests and validating code behavior from QA agents
  • Prototyping code solutions before committing from engineering agents

Not For

  • Sandboxed execution in untrusted environments (local execution has host access)
  • Long-running processes (designed for quick code snippets)
  • Production code deployment (local development/testing only)

Interface

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

Authentication

Methods: none
OAuth: No Scopes: No

No authentication — local code execution with host OS permissions. Security is controlled entirely by the agent trust model and OS-level isolation.

Pricing

Model: free
Free tier: Yes
Requires CC: No

Free and open source. Local execution only — no external costs. Language runtimes must be separately installed.

Agent Metadata

Pagination
none
Idempotent
Partial
Retry Guidance
Not documented

Known Gotchas

  • CRITICAL SECURITY: Code runs with full host OS permissions — only use with trusted agents
  • Language runtimes must be installed separately (Python, Node.js, Go, etc.)
  • No sandboxing — malicious or buggy code can affect the host system
  • Timeout handling for long-running code may vary — implement timeout awareness
  • Working directory for code execution may differ from expectations
  • Dependencies/imports not pre-installed — code requiring external packages may fail

Alternatives

Full Evaluation Report

Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for Code Runner MCP Server.

$99

Scores are editorial opinions as of 2026-03-06.

5210
Packages Evaluated
26151
Need Evaluation
173
Need Re-evaluation
Community Powered