More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    AgentGrid — MCP Security & Governance

    1

    Govern MCP tool execution with policy checks, approvals, audit trails, and safer defaults.

    Free

    1 useSecurity scanned
    AgentGrid — MCP Security & Governance

    AgentGrid — MCP Security & Governance

    Example session with this skill installed

    I need to deploy this application. Before running any shell commands or modifying infrastructure, evaluate the required actions through AgentGrid. Block anything prohibited by policy and request my approval before executing sensitive operations.

    • Read your context and instructions
    • Compiled the agentgrid

    AgentGrid Security Review
    AgentGrid evaluated the proposed deployment workflow before execution.
    Action Risk Decision
    Read deployment configuration Low ALLOW
    Run project tests Low ALLOW
    Build Docker image Medium ALLOW
    Access production credentials High REQUIRE_APPROVAL
    Modify production infrastructure High REQUIRE_APPROVAL
    Delete existing resources Critical BLOCK

    Decision
    Safe read-only and build operations may continue automatically.
    Access to production credentials and infrastructure changes requires explicit human approval.
    Destructive resource deletion has been blocked by policy.
    Next step
    Review the high-risk operations and explicitly approve them before AgentGrid permits execution.

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

    About this skill

    AgentGrid helps AI coding agents use MCP tools more safely by introducing a governance layer around tool execution. Instead of allowing an agent to invoke sensitive tools without scrutiny, AgentGrid provides a structured workflow for evaluating actions before execution. It can classify tool calls, apply configurable policies, require human approval for sensitive operations, block disallowed actions, and preserve an audit trail of execution decisions. The skill is designed for developers working with agentic systems, MCP servers, coding agents, and automation workflows where unrestricted tool access creates unnecessary risk. What AgentGrid helps with

    • Evaluate MCP tool calls before execution
    • Apply allow, block, or approval-required policies
    • Reduce accidental destructive or privileged actions
    • Introduce human approval for sensitive operations
    • Keep auditable records of security decisions
    • Encourage least-privilege and default-deny workflows
    • Make agent behavior easier to inspect and govern Typical use cases
    • AI coding agents with shell or filesystem access
    • Internal developer agents
    • MCP servers exposing sensitive tools
    • Infrastructure automation
    • Security-sensitive agent workflows
    • Teams evaluating AI-agent governance AgentGrid is intended to complement MCP infrastructure rather than replace it. The skill guides compatible agents through the governance workflow while AgentGrid provides the underlying policy and control layer.

    Requires AgentGrid control plane for technical enforcement. Does not replace OS permissions. Agent-side guidance only.

    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 any payment.

    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

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Requires

    GitHub
    AgentGrid MCP
    AgentGrid control plane

    Permissions

    Network Access
    Environment Variables
    Write Files
    Terminal / Shell
    Read Files

    Allowed Hosts

    api.github.com

    File Scopes

    src/**
    config/**
    *.md

    AgentGrid may inspect tool requests and relevant execution context in order to evaluate them against configured security policies. Network access is used to communicate with the AgentGrid governance layer. Sensitive actions should remain subject to explicit policy or human approval.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    No. AgentGrid is designed to work alongside MCP infrastructure. Your existing tools remain available while AgentGrid adds policy enforcement, approval workflows, and auditability around their use.

    No. The skill provides the agent-facing workflow and instructions. Enforcement features require the AgentGrid control plane or MCP integration.

    The skill uses the open SKILL.md format and is intended for compatible agent environments including Claude Code, Codex CLI, Cursor, VS Code Copilot, Gemini CLI, and other compatible tools.

    Developers and teams building AI agents that interact with sensitive tools, local environments, infrastructure, internal APIs, or production systems.

    New in Security