Skill UI

Guides creation of Cursor-specific subagents with isolated contexts, structured prompts, and frontmatter so multi-step workflows, verifiers, or specialists follow Cursor’s directories and delegation rules.

Guides Claude Code skill authors through designing and structuring workflow-based skills with numbered phases, decision routing, subagent delegation, and progressive disclosure so multi-step automations remain predictable and auditable.

Master the creation of complex, event-driven, and scheduled automation workflows. This guide covers setting up various triggers (email, chat, webhook, calendar), implementing conditional filters, managing multi-step processes, and orchestrating specialized AI agents in a multi-agent 'Society' for advanced delegation and collaboration.

This skill provides comprehensive reference architectures for integrating Lindy AI agents into production applications. It covers various complex patterns, including simple webhook callbacks, sophisticated event-driven pipelines, multi-agent delegation systems, scheduled workflows, and RAG-powered customer chat interfaces. Essential for designing and building robust, scalable, and reliable AI-powered systems.

A comprehensive guide for founders and first-time CEOs transitioning from technical expert to executive leader. It covers essential topics like identifying your founder archetype, mastering advanced delegation frameworks, managing personal energy (not just time), mitigating burnout, and planning for company succession. Use this when you feel like the bottleneck, struggle to scale, or need to build a strong executive team.

This skill utilizes the Microsoft Graph API to parse Office 365 Unified Audit Logs for security incident investigation. It is designed to detect critical indicators of account compromise, including the creation of suspicious email forwarding rules, unauthorized mailbox delegation changes, suspicious OAuth application consent grants, and other signs related to Business Email Compromise (BEC).

Details techniques for exploiting misconfigured Kerberos Constrained Delegation (KCD) in Active Directory. By abusing the S4U2Self and S4U2Proxy protocols, attackers can impersonate highly privileged users (e.g., Domain Admins) to achieve lateral movement or full domain compromise during authorized red-team engagements.

This skill provides a structured methodology for threat hunting DNS-based persistence mechanisms. It detects unauthorized changes such as DNS hijacking, dangling CNAME records enabling subdomain takeover, wildcard DNS abuse, or unauthorized NS delegation modifications. By utilizing passive DNS history (e.g., SecurityTrails API) and auditing cloud DNS logs (Route53, Azure, Cloudflare), users can identify persistent attack vectors that survive traditional remediation efforts.

This comprehensive guide details the methodology for Active Directory (AD) penetration testing. It covers deep domain enumeration, utilizing tools like BloodHound and Impacket, and executing advanced attacks such as Kerberoasting, AS-REP Roasting, and privilege escalation via delegation abuse. Use this when conducting authorized security assessments to discover and exploit misconfigurations to achieve domain compromise.

Lambda is a compact, machine-native communication protocol specifically designed for agent-to-agent (A2A) messaging. It functions as a shared, unambiguous vocabulary built from 'atoms' (2-character codes), completely bypassing the need for human natural language interpretation. It is ideal for orchestrators, task delegation, and logging structured coordination signals (like heartbeats or state changes) in multi-agent systems. Crucially, Lambda is intended only for machine communication, not for user-facing dialogue.

A robust framework for building complex, long-horizon AI agents using LangGraph. It addresses common pitfalls in advanced agent design, such as unbounded state growth (Virtual FS cleanup) and subagent persona leakage (explicit message overriding). Ideal for tasks requiring multi-step planning, task delegation, and self-correction through reflection loops.

Cavecrew provides a set of specialized subagents (investigator, builder, reviewer) designed for efficient code development delegation. Unlike vanilla agents, Cavecrew compresses output results, drastically reducing the token usage injected back into the main context. This allows users to maintain context over much longer, complex coding sessions, effectively preventing context exhaustion during large-scale refactoring, debugging, or multi-step feature development.