Skill UI

Deploys AI and NLP-powered detection to spot business email compromise by analyzing writing style, behavioral context, and intent, then integrates baselines, transformer models, and response workflows to surface impersonation attempts that rule-based filters miss.

Uses Context7 MCP to fetch up-to-date library or framework documentation for setup questions, API references, and code snippets, ensuring responses rely on live official sources rather than outdated training data.

Identify unauthorized DCSync credential theft by parsing Event ID 4662 for DS-Replication-Get-Changes requests from non-domain-controller accounts, correlating with SIEM and network context to validate and respond.

Uses the Kubernetes Python client to inspect pod security contexts, privileged mode, capabilities, namespace sharing, and hostPath mounts to highlight container escape vectors during security audits or incident response.

Handles IMO/Putnam/USAMO/AIME style problems by generating refined proofs, then running a fresh-context adversarial verifier to catch flaws, finally producing honest confidence reports and verified PDFs when possible.

Helps developers answer codebase questions by leveraging the `.understand-anything/knowledge-graph.json` graph; walks through project metadata, nodes, edges, and layers to explain dependencies and contexts.

Provide step-by-step explanations of specific files, functions, or modules, leveraging the understand-explain workflow to map graph context, layer placement, internal structure, and data flow for developers needing deep clarity.

An advanced debugging skill designed for AI agents. It utilizes gdb-cli to provide intelligent assistance for C/C++ programs, enabling the analysis of core dumps, debugging live processes, and investigating complex issues like crashes, deadlocks, and memory corruption. It correlates runtime state inspection with source code context for deeper insights.

This skill systematically audits a given project by analyzing its historical memory, rollout summaries, and local skill definitions. It identifies repeated complex workflows, recurring operational patterns, or areas where existing skills are stale. The audit process guides the user to recommend the highest-value skills to either add (if a new pattern emerges) or update (if current skills are outdated or incomplete), ensuring the AI agent operates with the most accurate and context-specific knowledge base for optimal project execution.

Blueprint turns a single objective into an executable, multi-step construction plan tailored for coding agents and multi-session teams, complete with context briefs, review gates, dependency awareness, and installable via `/blueprint <project> <objective>`.

A powerful utility designed to persist and manage the state and memory across long AI coding sessions. It automatically loads project context at startup, tracks activity using Git, and allows users to structuredly save session summaries, decisions, and next steps. Essential for maintaining continuity and focus in complex development workflows.

A rigorous, multi-agent framework designed for high-stakes quality assurance. It mandates that an output must pass independent review from two separate, context-isolated agents before being finalized. This method is crucial for minimizing hallucination, ensuring regulatory compliance, and maintaining factual accuracy when generating content for deployment or public consumption.