More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Castov MCP DevSecOps

    1

    Local DevSecOps tools for AI agents to scan secrets, calculate CIDR, and audit JWTs with zero data leaks.

    Free

    Security scanned
    Castov MCP DevSecOps

    Castov MCP DevSecOps

    Example session with this skill installed

    Scan this for secrets and tell me the usable host range for 10.0.0.0/24.
    STRIPE_KEY=sk_live_51P...
    AWS_SECRET=wJalrX...

    • Read your context and instructions
    • Compiled the castov mcp devsecops

    I have performed a local security scan and network calculation. The operations were executed entirely on your machine.

    DevSecOps Audit Results

    CategoryItemResult
    Secret ScanStripe Live KeyDetected (Redacted: sk_live_***)
    Secret ScanAWS Secret Access KeyDetected (Redacted: wJalrX***)
    NetworkCIDR Block10.0.0.0/24
    NetworkNetmask255.255.255.0
    NetworkFirst Usable IP10.0.0.1
    NetworkLast Usable IP10.0.0.254
    NetworkTotal Hosts254
    NetworkBroadcast10.0.0.255

    Next steps

    • Move the detected keys from your code into a secure vault or encrypted .env file.
    • Verify that your VPC route table includes the 10.0.0.0/24 range for the intended subnet.
    • Run castov_port if you plan to bind a service to a specific IP in this range.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    AI agents often hallucinate complex regex, network subnets, and cryptographic calculations, leading to security vulnerabilities and broken deployments. This skill provides 13 deterministic, local-first tools that move sensitive computations off the LLM and onto your local hardware.

    What it does

    • Secret Scanning audits code and logs for leaked AWS keys, Stripe tokens, and private PEM files using in-memory detection.
    • Network Calculation computes IPv4 CIDR ranges, subnet masks, and usable host capacities for VPC and firewall configuration.
    • JWT Auditing decodes tokens locally to inspect headers, payloads, and expiration claims without sending data to external services.
    • Cron Prediction parses POSIX expressions into human-readable text and predicts future execution times across multiple timezones.
    • Crypto Operations generates high-entropy secrets and computes SHA/MD5 digests using local Node.js crypto libraries.
    • Data Conversion transforms JSON to TypeScript interfaces and decodes timestamps from UUIDv7, ULID, and Snowflake IDs.

    How it works

    1. Configuration adds the Castov MCP server to your mcp.json or claude_desktop_config.json via npx.
    2. Detection identifies when a prompt involves networking, security tokens, or complex scheduling.
    3. Execution calls the specific local tool via stdio to perform deterministic logic.
    4. Verification returns structured JSON results to the agent for final presentation.

    Frameworks & tools

    Built for the Model Context Protocol (MCP). Integrates with Claude Desktop, Cursor, Windsurf, and VS Code Copilot. Uses Node.js crypto and local system utilities for port inspection.

    Why this beats prompting it yourself

    LLMs frequently fail at precise CIDR math and cron DST offsets. By using compiled local algorithms, you eliminate hallucinated network addresses and ensure that sensitive tokens never leave your machine during the audit process.

    Use cases

    • Auditing local .env files for secrets before a git commit.
    • Planning Kubernetes cluster IP ranges and VPC subnetting.
    • Debugging EADDRINUSE errors by identifying processes on specific ports.
    • Verifying JWT expiration and signature algorithms during API development.

    Known limitations

    Supports IPv4 only for networking tools. Secret scanning is limited to the provided content string and does not crawl the entire filesystem automatically.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.

    IPv4 only. Secret scanning is limited to the input string. Requires Node.js environment for npx execution.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 7 days ago

    • Free with an Agensi account

    Accounts
    Castov

    Permissions

    Terminal / Shell
    Network Access

    Allowed Hosts

    castov.com

    File Scopes

    **/*.json
    .env*
    src/**

    Castov MCP operates 100% locally via stdio JSON-RPC with zero-data logging, zero telemetry, and zero cloud dependencies. No external API keys or third-party account connections are required. All secret scanning, JWT decoding, and cryptographic operations execute entirely inside the user's local environment.

    Listed7 days ago

    What's inside

    Frequently Asked Questions

    The skill includes 13 deterministic tools that run locally on your hardware, ensuring that sensitive data like JWTs and private keys are never sent to an external LLM for processing. This eliminates the risk of data leaks while providing 100% accuracy on complex math like CIDR ranges and cron schedules.

    This skill is built for the Model Context Protocol (MCP) and is fully compatible with Claude Desktop, Cursor, Windsurf, and VS Code Copilot. It integrates seamlessly into any environment that supports standard MCP server configurations.

    Setup is simple: add the Castov MCP server to your local configuration file (such as `mcp.json`) using the provided npx command. Once configured, your AI agent will automatically detect and call these tools via stdio whenever you ask for networking, security, or conversion tasks.

    LLMs often hallucinate when calculating subnet masks, predicting cron execution times, or identifying specific character patterns in secrets. This skill replaces probabilistic AI guesses with compiled, deterministic algorithms, ensuring your VPCs and schedules are technically sound and error-free.

    You receive the full Castov MCP DevSecOps server package, including all 13 local-first tools for networking, cryptography, and security auditing. You also get lifetime access to all future updates and version releases within the major version cycle.

    Currently, the networking tools are optimized for IPv4 CIDR and subnet calculations. The secret scanner is designed to audit specific files or strings you provide to the agent rather than performing an unprompted, full-disk crawl of your entire filesystem.

    New in LLM Ops

    Free