enterprise-vpn-attack is a skill for Claude that organizes the external review of remote-access appliances: Cisco ASA and AnyConnect, Fortinet FortiGate, Citrix NetScaler, Palo Alto GlobalProtect, Pulse Secure and Ivanti Connect Secure, SonicWall and F5 Big-IP. It lists the URL paths, cookies and banners that reveal which product sits behind a login page.
It then offers a table of known pre-authentication and authentication-bypass advisories from 2018 to 2026, notes on SAML service-provider settings, default-account and lockout cautions, ways to identify the authentication backend, a short probe sequence and a list of anti-patterns. It belongs to the external-engagement part of the bundle.
What You Get
- Vendor fingerprints for seven families of SSL VPN products, based on paths, cookies and banners
- A CVE matrix from 2018 to 2026 organized by vendor
- SAML configuration checks and notes on identifying the authentication backend
- A short probe sequence, a list of nuclei templates for triage and operational discipline notes
- Anti-patterns and links to neighboring skills
Common Use Cases
- Working out which VPN product and version an organization exposes during external recon
- Checking an authorized client’s gateways against a list of known advisories before a report
- Reviewing SAML and authentication backend settings on a remote-access appliance in scope
What Makes It Unique
It treats the VPN appliance as its own surface with vendor-specific signals. It states its boundaries up front: no internal lateral movement, no client-side VPN bugs and no IPsec or OpenVPN work, and it advises care around lockouts when default accounts are tried.
Under The Hood
The skill is a single SKILL.md of instructions with no bundled scripts. Following it, the agent sends HTTP requests with curl to the gateway hosts named in scope to read headers, login pages and version hints, and it can run nuclei templates against them. It also covers testing a small set of default accounts, with a caution about lockouts. Notes and results may be written to files in the working folder. The file mentions credentials only as things to avoid triggering lockouts with, and cites historic credential-dump formats as background. Use it only against gateways you are authorized to test.