Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    ast microsoft

    1

    Investigate and respond to threats across Microsoft 365, Entra ID, Azure, and Defender using MCP and CLI tools.

    $199

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    ast microsoft

    ast microsoft

    Example session with this skill installed

    Check Entra ID for any risky users and list the most recent Defender alerts.

    • Read your context and instructions
    • Compiled the ast microsoft

    I have completed the investigation into risky users in Entra ID and gathered the latest alerts from Microsoft Defender.

    Security Investigation Report

    SourceIdentifierRisk Level / SeverityDescriptionTimestamp (UTC)
    Entra IDuser.alpha@company.comHighLeaked Credentials2023-10-27 14:22:01
    Entra IDuser.beta@company.comMediumUnfamiliar sign-in properties2023-10-27 15:45:10
    DefenderAlert-8821InformationalManual investigation started2023-10-27 16:00:12
    DefenderAlert-8825HighPossible Cobalt Strike activity2023-10-27 16:15:45

    Gaps

    • The search was limited to the last 24 hours.
    • No errors were reported by the az-search tool.
    • Results cover Entra ID and Defender only; Sentinel KQL hunting was not performed.

    Next steps

    • Needs your approval: Revoke sessions for user.alpha@company.com to mitigate the leaked credential risk.
    • Needs your approval: Isolate the host associated with Defender alert Alert-8825.
    • Run a Sentinel KQL query to check for lateral movement from these accounts.

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

    About this skill

    Securing and managing Microsoft 365 or Azure environments often requires jumping between multiple portals and running disjointed PowerShell scripts. This skill provides a unified command-line and MCP interface to investigate and respond to threats across the entire Microsoft stack from your agent.

    What it does

    • Security Investigation queries Entra ID, Defender alerts, and Sentinel incidents using KQL to hunt for threats.
    • Incident Response disables compromised accounts, revokes sessions, and contains hosts across on-prem AD and Entra.
    • Email Defense searches mailboxes, analyzes headers, and purges phishing emails or releases items from quarantine.
    • Device & App Management audits Intune device status, manages the Win32 app catalog, and tracks install status.
    • Compliance & Governance manages Purview eDiscovery cases, litigation holds, and Teams federation blocks.
    • Resource Optimization analyzes Azure cost data and identifies VM rightsizing opportunities.

    How it works

    1. Prerequisites requires the user to create read, write, and compliance app registrations in the Entra portal.
    2. Installation involves running an install.sh script to set up a local Python virtual environment and offline AST packages.
    3. Configuration uses a .env file or system keychain to store API credentials securely.
    4. Execution leverages specialized CLIs like az-search, az-exch, and az-sentinel or their corresponding MCP tools.

    Frameworks & tools

    Built for Python 3.12+, utilizing Microsoft Graph API, Azure Management API, and Sentinel KQL. Certain Exchange and Purview features require PowerShell 7 and the ExchangeOnlineManagement module.

    Why this beats prompting it yourself

    Generic prompts often hallucinate PowerShell syntax or overlook the complex API permissions required for M365 security actions. This skill provides tested, specific tooling that handles the nuances of Microsoft's various API endpoints while enforcing a strict safety model for destructive actions.

    Use cases

    • Disable a compromised user account and revoke all active sessions across Entra and on-prem AD.
    • Search all mailboxes for a specific phishing SHA256 hash and purge the messages.
    • Query Sentinel for high-severity incidents and pull associated Defender hunting telemetry.
    • Audit Azure tenant spend and identify underutilized VMs for cost reduction.

    Known limitations

    Does not support Okta, Google Workspace, or sovereign clouds like GCC High and DoD. Write operations require manual environment variable gates to be enabled for MCP access.

    No support for Okta, Google Workspace, or GCC High. Requires manual Entra app registration. Certain Exchange tasks need PowerShell 7. Writes are gated by env vars.

    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

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 24 days ago

    Requires

    Python 3.12+
    macOS
    Linux
    Bash 3.2+
    pip
    venv
    requests
    python-dotenv
    platformdirs
    keyring
    fastmcp
    msal 1.24+
    ldap3 2.9+
    Python
    Windows

    Permissions

    Terminal / Shell
    Browser
    Network Access
    Read Files
    Write Files
    Environment Variables

    File Scopes

    ~/.local/share/ast/venv/**
    ~/.config/ast/.env
    ~/.claude/skills/ast-microsoft/**
    ./.env
    (read
    if
    present)
    ~/Library/Application
    Support/<tool>/.env
    and
    ~/.config/<tool>/.env
    per-tool
    fallback)
    ~/.cache/agentic-security-toolkit/**
    (pending
    dry-run
    /
    exec
    tokens)
    *.json
    *.csv
    at
    a
    path
    you
    choose
    (--output
    sectools_export)
    Support/az-cal/
    teams-edit/
    (macOS)
    ~/.config/az-cal/
    ~/.config/teams-edit/
    (Linux):
    delegated_token_cache.json
    Support/agentic-security-toolkit/{exch-edit
    intune-edit}/**
    or
    ~/.local/state/agentic-security-toolkit/...
    change
    snapshots
    for
    rollback
    Certificate
    PEM
    files
    point
    AZURE_*_CERT_PATH
    (read)
    Download/export
    paths
    (drive-get
    --out-dir
    Teams
    conversation
    export)

    Terminal/Shell runs `install.sh`, the 12 CLIs, `sectools` and, for some Exchange/Purview commands, `pwsh`; Read/Write Files cover the venv, your `.env`, certificate files, delegated token caches, change snapshots and exports you request; Environment Variables carry the tenant, app and AD credentials and the gates (AZURE_* are read from env/.env only, never the keychain); Network Access reaches PyPI at install and Microsoft's commercial-cloud endpoints (and LDAPS to your DCs) at run time. Reads use the read app. Writes use separate write / compliance apps, ask [y/N] on the CLI, and over MCP each area needs its own gate set to exactly 1 (Exchange and Intune CLI writes also require their gate). The skill requires your explicit approval of each change.

    Listed1 month ago
    Updated24 days ago

    What's inside

    Frequently Asked Questions

    This skill consolidates Microsoft 365, Entra ID, Azure, and Defender management into a single interface, eliminating the need to switch between multiple portals or manually debug complex KQL and PowerShell scripts for incident response.

    The skill is designed for Python 3.12+ environments and integrates seamlessly with agents supporting the Model Context Protocol (MCP) or standard CLI execution, such as those running on local workstations or secure administrative hosts.

    Yes, for security reasons, destructive or 'write' operations like disabling accounts or purging emails require manual environment variable gates to be enabled, ensuring your agent doesn't perform sensitive actions without explicit permission.

    Setup involves creating specific Entra ID app registrations for API access, followed by running a provided installation script that configures a local Python environment and the necessary specialized CLI tools.

    The package includes the core Python codebase, the specialized CLI tools (az-search, az-exch, az-sentinel), the MCP integration layer, and detailed documentation for setting up the required Azure permissions.

    While it provides comprehensive coverage for standard Microsoft 365 and Azure global tenants, it currently does not support sovereign clouds such as GCC High, DoD, or third-party identity providers like Okta.

    New in Agents & Orchestration