More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Email Security: Phishing and Header Analysis

    1

    Analyze a suspicious email like a responder: safe triage, SPF/DKIM/DMARC alignment checks, URL and attachment dissection, BEC pattern recognition, IOCs, and tenant hardening.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Email Security: Phishing and Header Analysis

    Email Security: Phishing and Header Analysis

    Example session with this skill installed

    A user forwarded me this from the gateway quarantine: it claims to be a Microsoft 365 password-expiry notice, From shows our own IT helpdesk address, and the button links to an oauth2/authorize URL. Analyze it and tell me what to do.

    • Read your context and instructions
    • Compiled the email security

    VERDICT: malicious — score 9/12. TRIAGE: worked from the gateway's raw copy; original preserved, SHA-256 recorded; nothing forwarded to users.

    HEADERS: Received chain breaks at hop 2, where a hosting-range IP claims to relay for the company domain. Authentication-Results: spf=pass — but for envelope domain secure-notify365.net; dkim=none; dmarc=fail. The SPF pass belongs to the attacker's domain, not the helpdesk From it is impersonating: pass without alignment. Reply-To points to an unrelated free-mail account. Scores: Authentication 2, Header integrity 2.

    URL: the button is an OAuth consent URL against the real Microsoft identity endpoint — the landing domain is legitimate, which is the trap: approving it grants a third-party app (registered 2 days ago, publisher unverified) mail.read and offline_access. This is consent phishing; no password is ever typed. URL 2, Content 1 (expiry urgency).

    ATTACHMENT: none. Infrastructure 2 (sender domain created 6 days ago).

    ACTIONS: purged tenant-wide — sibling sweep found 14 recipients, 2 had clicked, 1 had consented; that app registration was revoked, sessions reset, and the user briefed by phone. IOCs filed: sender domain, sending IP, app ID, subject pattern. Detection gap: user consent to third-party apps was unrestricted. Hardening action: admin-consent workflow enabled, DMARC moved from p=none to p=reject with aggregate reporting.

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

    About this skill

    Most phishing "analysis" is a glance at the rendered message and a gut call — exactly the parts of the email the attacker controls. This skill runs the sample the way an incident responder does. Triage happens on the raw source with the original preserved and hashed, never forwarded to real users. The header walk reads the Received chain bottom-up and asks the question that decides most cases: not whether SPF or DKIM passed, but whether the authenticated domain aligns with the visible From — a pass for the attacker's own domain proves nothing. URLs are compared landing-domain vs. claimed brand (including OAuth consent phishing), attachments are typed by magic bytes and detonated only in a sandbox, and BEC patterns get their own pass because they carry persuasion instead of malware. Every sample is scored on a six-signal rubric to a verdict, and the analysis is not finished until it produces detection-ready IOCs, a named detection gap, and at least one tenant hardening action.

    • It needs the raw message source; a screenshot or forwarded render strips the headers the verdict depends on.
    • It does not replace your email gateway or SIEM — it produces the analysis and IOCs those tools consume.
    • Compromised legitimate accounts pass authentication by definition; the skill flags that pivot, but containment is an identity-team action.

    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

    Permissions

    Terminal / Shell
    Listed24 days ago
    Updated21 days ago

    What's inside

    Frequently Asked Questions

    No — SPF passing only proves the sending IP was authorized for the envelope domain, which may be the attacker's own domain. The skill checks alignment: whether the authenticated domain matches the visible From domain. A pass without alignment is the classic spoof signature.

    Detonation belongs in a sandbox or a disposable VM that is not your workstation and not on the corporate network — and only after static checks (real file type vs. extension, macro presence, embedded URLs). The skill sequences static analysis first so most attachments never need detonating at all.

    The verdict is the smallest part of the output. The same run produces IOCs that block the campaign's next hundred sends, a sibling sweep that finds the other recipients before they click, and the tenant fix (DMARC enforcement, consent restrictions) that stops the repeat.

    New in Security