mcpField
D
DevOps & Cloud Code Auditor

DevOps & Cloud Code Auditor

Remote
neoninnovationlab.comDeveloper Tools0Updated 6w ago

About DevOps & Cloud Code Auditor

DevOps & Cloud Code Auditor is a static-analysis MCP server for engineering teams, checking Terraform, PostgreSQL migrations, and Next.js code for cost, locking, and security issues. It analyzes code without executing it or connecting to databases.

Key features

  1. 01Detect cloud cost issues in Terraform
  2. 02Find PostgreSQL migration lock hazards
  3. 03Check Next.js API security rules
  4. 04Audit snippets, Git repositories, or local directories
  5. 05Return structured findings and fix code

Use cases

  1. 01Check Terraform configuration before deployment
  2. 02Review PostgreSQL migrations before production
  3. 03Inspect Next.js APIs and Server Actions

MCP support

The MCP server provides tools to audit repositories, code snippets, and local directories, returning structured findings, severity information, and fix code. Analysis is static and does not execute code or connect to databases.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://neon_innovation_lab--devops-code-auditor.apify.actor/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

DevOps & Cloud Code Auditor (MCP Server)

Model Context Protocol Glama Quality Score Python 3.12 License: MIT Transport: stdio

A production-ready Model Context Protocol (MCP) Server providing AI coding assistants (Claude Desktop, Cursor, Windsurf, Cline) with automated static analysis tools to prevent production outages, cloud cost leaks, and security vulnerabilities.


🌟 Capabilities & Audited Rule Sets

This server inspects codebases across three mission-critical engineering vectors:

  1. Cloud Cost Leaks (Terraform / HCL / AWS):

    • EBS_GP2: Identifies legacy gp2 EBS volumes (gp3 provides equal durability with ~20% baseline cost reduction and higher baseline IOPS).
    • PUBLIC_IP: Flags unneeded associate_public_ip_address = true incurring hourly AWS IPv4 charges ($0.005/hr/IP).
    • CW_NO_RETENTION: Catches aws_cloudwatch_log_group missing retention policies (preventing infinite unbudgeted log ingestion costs).
    • S3_NO_LIFECYCLE: Detects S3 buckets lacking Glacier lifecycle transitions.
    • RDS_MULTIAZ_NONPROD & RDS_PROVISIONED_IOPS_NONPROD: Flags doubled database infrastructure costs on development and staging environments.
  2. PostgreSQL Migration Table Lock Hazards:

    • INDEX_NOT_CONCURRENT: Detects CREATE INDEX missing CONCURRENTLY (which acquires an EXCLUSIVE lock and halts table writes in production).
    • NOT_NULL_VOLATILE_DEFAULT: Catches ADD COLUMN ... NOT NULL DEFAULT <func()> rewriting every table row under an ACCESS EXCLUSIVE lock.
    • FK_NO_INDEX: Flags unindexed foreign key constraints causing cascading table-level sequential lock scans on parent updates/deletes.
    • FOR_UPDATE_UNBOUNDED: Catches SELECT ... FOR UPDATE missing LIMIT or SKIP LOCKED, preventing deadlocks.
  3. OWASP & Full-Stack Next.js API Security:

    • NEXT_PUBLIC_SECRET_LEAK: Flags sensitive tokens and private API keys prefixed with NEXT_PUBLIC_ that leak into public client browser bundles.
    • NEXTJS_SERVER_ACTION_NO_AUTH: Flags Next.js 14/15 Server Actions exporting mutations without explicit session auth verification.
    • SQLI_RAW_INTERPOLATION: Identifies raw SQL query templates using string interpolation instead of parameterized inputs.
    • SSRF_UNVALIDATED_FETCH: Catches external fetch calls consuming user inputs without protocol/hostname validation.

🛠️ MCP Tools Exposed

This server implements the official Model Context Protocol specification (tools/list and tools/call):

1. audit_devops_repository

Clones a public Git repository shallowly without executing hooks, performs AST and regex static syntax audits, and returns structured findings and a formatted Markdown report.

  • Parameters:
    • repository_url (string, required): Public Git URL of the repository to audit (e.g. https://github.com/example/cloud-infra).
    • sub_directory (string, optional): Specific subdirectory within the repository to inspect (e.g. terraform/ or prisma/migrations/).
  • Returns:
    • status: "success" | "error"
    • total_findings: Number of detected issues
    • findings: Array of detailed vulnerability records (file, line number, rule ID, severity, message, fix code)
    • markdown_report: Human-readable summary table
2. audit_devops_code

Directly audits a raw code snippet of Terraform (HCL), PostgreSQL migration (SQL), or Next.js API/Server Action code.

  • Parameters:
    • code (string, required): The source code string to audit.
    • file_type (string, optional, default: "auto"): One of "terraform", "postgres", "nextjs", or "auto".
  • Returns: Structured JSON findings with exact drop-in code fixes.
3. audit_local_devops_directory

Audits a local filesystem directory on the host machine before submitting a Pull Request or deploying to staging.

  • Parameters:
    • directory_path (string, required): Absolute filesystem path to the target folder.
  • Returns: Structured JSON findings and severity breakdown.

🚀 Client Configuration & Quickstart

1. Claude Desktop Setup

Add the server to your claude_desktop_config.json:

{
  "mcpServers": {
    "devops-code-auditor": {
      "command": "python3",
      "args": ["/path/to/devops-code-auditor/server.py"]
    }
  }
}
2. Cursor IDE / Windsurf Setup

In Cursor, navigate to Settings → Features → MCP → Add New MCP Server:

  • Name: devops-code-auditor
  • Type: command
  • Command: python3 /absolute/path/to/devops-code-auditor/server.py
3. Docker Execution

Run the isolated container via Docker:

{
  "mcpServers": {
    "devops-code-auditor": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/ansarii/devops-code-auditor"
      ]
    }
  }
}

💻 Example Agent Usage

Once connected, your AI assistant can evaluate migration files and cloud manifests on demand:

User: "Check this PostgreSQL migration before I run it on production:
CREATE INDEX idx_orders_user_id ON orders (user_id);"

Claude (Tool Call):
audit_devops_code({
  "code": "CREATE INDEX idx_orders_user_id ON orders (user_id);",
  "file_type": "postgres"
})

Result:
{
  "status": "success",
  "total_findings": 1,
  "findings": [
    {
      "suite": "postgres-locks",
      "rule": "INDEX_NOT_CONCURRENT",
      "severity": "high",
      "file": "migration.sql",
      "line": 1,
      "message": "CREATE INDEX on \"orders\" without CONCURRENTLY blocks table writes during build.",
      "fix": "CREATE INDEX CONCURRENTLY idx_orders_user_id ON orders (user_id);"
    }
  ]
}

🔒 Security Guarantee: Zero Execution

This server executes 100% static analysis using AST pattern matching and regex verification. It never executes untrusted shell commands, arbitrary Python code, or database connections on your machine.


📜 License

MIT License — Copyright (c) 2026 Neon Innovation Lab. Maintained by Neon Innovation Lab.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.