Skills
96,770 packaged capabilities and playbooks for agents. Install any of them with one command.
Security
Hunt MFA / 2FA bypasses across 7 patterns: (1) MFA not enforced on sensitive endpoints (password changes or email-change acceptance without an MFA challenge), (2) skipping the MFA step by navigating directly to a post-login URL, (3) MFA-token replay (the same code is accepted twice), (4) brute-forcing the 6-digit OTP without rate limiting (10^6 attempts at server speed), (5) race conditions in OTP validation, (6) recovery-code exposure via /api/me, and (7) backup-factor downgrade (SMS factor without rate limiting). Also covers the chain: cookie theft + password oracle + no step-up = ATO without an MFA challenge. Detection: trace the auth flow in Burp, find every state transition, check whether MFA is middleware-gated or enforced per endpoint, and check OTP entropy and rate limits on OTP validation. Validate that an attacker session can reach the post-MFA state. Use when hunting auth bypasses, MFA flows, or chaining primitives toward ATO.
Hunt session management vulnerabilities — session fixation (no regeneration on login), insufficient invalidation on logout / password-change / email-change, predictable or low-entropy session IDs, JWT-as-session with no exp/revocation, refresh-token rotation/reuse-detection gaps, OAuth/SSO session linkage, device-bound-session (DBSC) downgrade, and cookie attribute issues (Secure/HttpOnly/SameSite/__Host-). Validate with TWO real sessions (attacker A + victim B), body-diff every 200, and OOB confirmation for theft chains. Medium to Critical (fixation→admin hijack, no-invalidation→persistent ATO).
Hunting skill for ssrf vulnerabilities. Built from 15 public bug bounty reports including AWS metadata SSRF (HackerOne $25k Analytics PDF, Shopify Exchange $25k, Capital One 106M-record breach, Dropbox/HelloSign $4,913), GCP metadata SSRF (Snapchat $4k), Azure IMDS SSRF (Azure DevOps $15k chain, ChatGPT Custom Actions MSRC), DNS rebinding SSRF (Concrete CMS, GitLab UrlBlocker), gopher-protocol-to-Redis-RCE (Yahoo Mail $15k), link-preview SSRF (Reddit Matrix $6k), and headless-browser PDF-generator SSRF chains. Use when hunting SSRF on any target — OOB Collaborator confirmation mandatory for blind cases.
Online password brute-forcing with Hydra. Use when you need to brute-force, password-spray, or credential-stuff SSH/FTP/HTTP/SMB/RDP/MySQL/MSSQL and other services. Hydra is the classic online brute-force tool, supporting 50+ protocols. Use this skill for any scenario involving online password brute-forcing, login cracking, or protocol brute-force testing
IOC (indicator of compromise) analysis and countermeasure methodology. Blue team perspective: how to collect, enrich, and correlate IOCs for threat hunting. Red team perspective: how to avoid generating identifiable IOCs from your own infrastructure and tools, and how to make IOCs expire quickly. Use when the red team needs to assess its own exposure or plan C2 infrastructure
IRify SAST — AI-powered static application security testing. Compile source code into SSA IR, then use SyntaxFlow DSL to trace data flow across function boundaries, detect vulnerabilities (SQLi, RCE, XXE), and answer questions like "where does user input go?". Supports 7 languages (Java, PHP, JS, Go, Python, C, Yak) and incremental compilation via ProgramOverLay. Requires yaklang MCP server: yak mcp -t ssa
Offline password cracking with John the Ripper. Use when you need to crack hashes (MD5/SHA/NTLM/Kerberos/ZIP/RAR/PDF/SSH Key, etc.). John supports automatic hash type detection, dictionary attacks, rule-based mangling, and incremental brute-forcing, and ships with the *2john toolchain for extracting hashes from various formats. Use this skill for any scenario involving offline password cracking, hash recovery, or password auditing
Traffic hijacking and sensitive information theft in Kubernetes sidecar containers. Use when a target Pod has an Istio/Envoy/Linkerd sidecar, when the challenge mentions an 'invisible bystander' or 'shared network', or when you need to sniff traffic from inside a Pod. Covers tcpdump packet capture, plaintext traffic capture from sidecars, and exploitation of shared network namespaces. Use this skill whenever you find signs of a sidecar or multiple co-existing containers in a K8s Pod
Kubernetes Admission Webhook abuse and policy engine exploitation. Use when the cluster has Kyverno/OPA Gatekeeper/custom webhooks, when DNS scanning reveals kyverno-svc or gatekeeper services, or when you need to extract injected secrets from a Mutating Webhook. Core technique: forging AdmissionReview requests. Use this skill for any scenario in K8s where a webhook service or policy engine is found
Kubernetes cluster DNS service discovery and reconnaissance with k8spider. Use when you need to enumerate Services in a K8s cluster with low privileges, probe DNS services, scan network ranges for PTR/SRV records, or attempt AXFR zone transfers. k8spider needs only DNS access to discover all services in a cluster, with no API Server permissions required. Use this skill for scenarios involving K8s service discovery, DNS reconnaissance, cluster information gathering, PTR scanning, or SRV enumeration. Trigger it also when the user mentions k8spider, K8s DNS enumeration, cluster service scanning, or DNS zone transfer
Complete attack methodology for Kerberoasting and AS-REP Roasting. Use when you hold any domain user credentials (Kerberoasting) or discover accounts that do not require pre-authentication (AS-REP Roasting). Covers SPN discovery, ticket requests, offline cracking, AES vs RC4 strategy selection
A collection of advanced Kerberos protocol attack techniques. Use the advanced techniques in this skill when basic Kerberoasting/AS-REP Roasting and delegation attacks are not enough. Covers sAMAccountName spoofing (noPac/CVE-2021-42278), Kerberos relay (KrbRelayUp), Timeroasting (unauthenticated hash extraction over NTP), ASREQroast (MITM AS-REQ capture), SPN-jacking (targeted Kerberoasting), UnPAC-the-hash (certificate → NT Hash), Sapphire ticket (highly stealthy ticket forgery), Bronze Bit (CVE-2020-17049 delegation bypass), and the RODC attack chain (key list export/RODC Golden Ticket/RODC DACL exploitation).
Log analysis and log evasion methodology. Understand how the blue team traces attacks through logs (SIEM/Event Log/Syslog), and how the red team avoids logging or precisely removes traces. Use when designing operationally quiet actions or analyzing log monitoring coverage
Security monitoring agent for Claude Skills and MCP servers. Real-time protection layer (PreToolUse hook, zero LLM cost by default): hard-blocks confirmed-malicious tool calls (known-bad domains from real incidents and an auto-updating URLhaus malware feed) and, for merely suspicious ones (credential exfiltration, reverse shells, curl|bash pipes, raw-IP URLs, cloud-metadata/IMDS, config/persistence writes), asks you to approve or deny at the native prompt instead of blocking outright. Approving a flagged path/domain is remembered so it stops asking (trust builds as you confirm what you use). v3 adds: multi-step attack-chain detection (credential access then egress), cross-server data-flow tracking, a config/MCP scanner + integrity baseline (catches a malicious hook planted in a cloned repo), a shadow/audit-only mode (SENTINEL_SHADOW: never blocks, just tallies what it would have stopped), and an OPTIONAL, off-by-default AI escalation layer (SENTINEL_AI: only for ambiguous cases, token-budgeted, hardened against prompt injection). Localized messages (Spanish/English). Threat data ships base64-at-rest so a host antivirus does not flag the install. The v1 static analysis still runs on demand: scans installed skills/MCPs against vulnerability databases (GitHub Advisory DB, vulnerablemcp.info, CVE feeds, mcpscan.ai, Snyk, ClawHub/VirusTotal) and community alerts, maintains a local threat database, performs coherence and update-diff analysis. Use this skill whenever: the user asks about security of their skills or MCPs, wants to audit installed plugins, enable real-time protection, mentions "vulnerability", "CVE", "malicious skill", "security scan", "threat", "audit", "runtime protection", "block", says "is this skill safe?", asks to check dependencies, or wants ongoing security monitoring. Also trigger proactively when the user is about to install a new skill or MCP server.
High-speed port scanning with naabu. Use when you need port discovery or liveness detection on target hosts/networks. naabu is a fast port scanner from ProjectDiscovery supporting SYN/CONNECT/UDP scans, with performance far exceeding nmap, plus batch targets, CDN exclusion, and nmap integration. Use this skill for any scenario involving port scanning, port discovery, host liveness detection, or network probing. If naabu results are insufficient, fall back to nmap
Web server vulnerability scanning with Nikto. Use when you need to detect known vulnerabilities, outdated software versions, dangerous files/CGI, or misconfigurations on a web server. Nikto includes 7000+ checks covering common OWASP issues. Use this skill for any scenario involving web vulnerability scanning, server security checks, or configuration auditing
Comprehensive penetration testing workflow using Kali Linux tools via MCP. Guides agents through reconnaissance, vulnerability assessment, exploitation, privilege escalation, and reporting. Use when conducting security assessments, CTF challenges, vulnerability testing, or red team exercises.
Use when building ReactJS web authentication with the Ping Orchestration JavaScript SDK (DaVinci) using `@forgerock/davinci-client` — scaffolds a complete React SPA authentication flow against PingOne DaVinci. Handles DaVinci client configuration, OIDC token exchange, dynamic DaVinci collector rendering in React (text, password, social login, MFA, FIDO2/passkeys, PingOne Protect), protected routes, user profile, and logout. Invoke when the user asks to "use @forgerock/davinci-client in React", "render DaVinci collectors in a React app", "set up PingOne DaVinci authentication in React", "build a React SPA with DaVinci", "add DaVinci login to my React app", or "use the DaVinci SDK with Vite and React". Also covers migrating a React app from `@forgerock/javascript-sdk` to `@forgerock/davinci-client`.
Guidance for Microsoft Purview Records Management — declaring, managing, and disposing records across SharePoint, OneDrive, Exchange, and Teams. Covers retention labels with record / regulatory record options, file plan import, event-based retention (employee leaves, contract expires), disposition review (single- and multi-stage), proof of deletion / records of disposition, retention label policies vs auto-apply policies (KQL/sensitive info types/trainable classifiers), label-aware DLP, integration with Information Governance vs Records Management licensing, and role separation between records managers and admins. WHEN: records management Purview, file plan, retention label record, regulatory record, event-based retention, disposition review, record declaration SharePoint, immutable records, audit-proof deletion, file plan import. DO NOT USE for non-records data lifecycle (use purview-data-lifecycle), DLP policies (use purview-dlp-policy), or eDiscovery (use purview-ediscovery).
Validate systems and processes against GDPR/CCPA privacy regulations, privacy-by-design principles, ADA/WCAG accessibility standards, data processing agreements (DPAs), and provide compliance checklists with regulatory change monitoring guidance.
Audit mobile app source code for insecure data storage vulnerabilities per MASVS-STORAGE controls and MASTG tests. Use when reviewing how an app stores sensitive data locally, checks for data leakage, or assesses data-at-rest protection on Android or iOS.
Security audit orchestrator - parallel dependency, SAST and auth/config agents consolidated into an OWASP-mapped severity report, plus PHP 8, Twig and Craft CMS hardening references. Triggers on: security review, security audit, OWASP, XSS, SQL injection, CSRF, authentication, authorization, secrets management, input validation, vulnerability scan, dependency audit, composer audit, PHP security, Twig escaping, template injection, Craft CMS security, devMode, allowAnonymous, security key, DDEV. Use when auditing or hardening an app before release - e.g. 'security review this PR', 'is this Twig |raw safe?', 'audit our Craft config for production', 'does this controller need allowAnonymous?', 'triage composer audit output', 'diff DDEV against production PHP settings'.
Load this skill before shipping a UI change, to statically review the diff for design-craft regressions (accessibility, motion, responsive, and visual or UX drift), correctness and security issues, exposed secrets or credentials, and leftover commented-out code, then write a commit message and pull request description that match the repository's conventions, and offer to commit and push. Triggers on "ship this", "review this", "ready to ship". This is the ship gate at the end of the design → build → ship pipeline.
Sliver C2 framework operations guide. Use when you need to establish command-and-control channels, manage implants, or perform post-exploitation. Supports mTLS/HTTPS/DNS/WireGuard multi-protocol, Session/Beacon dual modes, SOCKS5 proxy, and multi-layer pivoting over TCP/Named Pipe. Suitable for red team operations and penetration testing scenarios.
Better Auth on TanStack Start: betterAuth() config, plugins, adapters, sessions, and FMC route protection (layout beforeLoad for route UX, endpointAuth/guards for data boundaries, headers, tanstackStartCookies plugin, OAuth sign-in). Use for auth.ts, Better Auth, sign-in flows, session helpers, or secured routes/APIs.
Tencent Cloud penetration testing methodology. Use when the target uses Tencent Cloud services, when cos.*.myqcloud.com assets are discovered, when Tencent Cloud SecretId/SecretKey is obtained, when the metadata at metadata.tencentyun.com is reachable from inside a CVM instance, or when a security assessment of Tencent Cloud services such as CAM/CVM/COS/TencentDB/TKE/SCF is required. Covers CAM privilege escalation, CVM takeover, COS object storage exploitation, TencentDB database attacks, TKE container clusters, SCF cloud functions, CLB load balancing, CLS log service, and KMS key management
Methodology for delivering tools and files to a compromised target host. Use when you already have a shell/webshell/RCE and need to transfer tools such as fscan, frp, chisel, linpeas, or mimikatz to the target for execution. Covers file transfer techniques across both Linux and Windows platforms (wget/curl/certutil/bitsadmin/PowerShell/python/nc/base64/SMB, etc.), including why tools should be executed on the target machine rather than locally, how to handle the absence of outbound internet access, how to bypass antivirus detection, and post-execution artifact cleanup. Use this skill for any scenario involving 'transferring files to the target', 'executing scanning tools on the target', 'internal network tool delivery', or 'file staging'
This skill should be used when the user asks to "identify web application vulnerabilities", "explain common security flaws", "understand vulnerability categories", "learn about injection attacks", "review access control weaknesses", "analyze API security issues", "assess security misconfigurations", "understand client-side vulnerabilities", "examine mobile and IoT security flaws", or "reference the OWASP-aligned vulnerability taxonomy". Use this skill to provide comprehensive vulnerability definitions, root causes, impacts, and mitigation strategies across all major web security categories.
Encrypting data before storing on Walrus using Seal (threshold encryption with onchain access control). Use when the user needs to store private or sensitive data on Walrus, implement access control for blob content, encrypt blobs before uploading, use @mysten/seal for threshold encryption, or understand Walrus data security guarantees (availability, integrity, confidentiality). Also use when the user asks about Nautilus (TEE-based off-chain computation) in the context of Walrus data. All blobs on Walrus are public by default.
Receive and verify WorkOS webhooks. Use when setting up WorkOS webhook handlers, debugging WorkOS-Signature verification, or handling enterprise auth events like dsync.user.created, dsync.group.user_added, connection.activated, user.created, or session.created.
Automated web vulnerability scanning with xray. Use when you need comprehensive vulnerability scanning of a web application (XSS/SQLi/command injection/SSRF/XXE/path traversal/file upload/weak credentials, etc.). xray is a comprehensive web security assessment tool by Chaitin Tech, supporting three modes: active scanning, passive proxy scanning, and basic crawler scanning, with rich built-in detection plugins and community POCs. Use this skill for any scenario involving xray vulnerability scanning, web security assessment, or passive proxy scanning
Attacks on domain trust relationships. Use when the target has a multi-domain or multi-forest environment. Covers parent-child domain privilege escalation (Golden Ticket + ExtraSid), cross-forest attacks (SID History/MSSQL Trust Links), and exploitation of one-way trusts. Load first when child-domain Domain Admin access has been obtained or a trust relationship is discovered.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.