Skill UI

Architecture and integration sketches for LLM applications, covering RAG retrieval, tool dispatch, agent loops, caching, privacy boundaries and verification requirements before production use.

Expert guidance for building production-grade AI agents with LangChain 0.1+ and LangGraph, covering architecture, memory, RAG, tools, deployment, and evaluation.

Conduct authorized security assessments for LLM applications and AI agents. Covers prompt injection, tool abuse, RAG exposure, memory poisoning, and system prompt extraction following OWASP LLM/ASI Top 10 guidelines. Includes red-teaming, vulnerability mapping, and compliance engineering for AI systems to ensure safety and robustness.

Loki Mode is a multi-agent autonomous startup system designed to execute workflows from PRD to production without human intervention. It integrates research-enhanced AI models including OpenAI SDK, DeepMind, Anthropic, and AWS Bedrock. This skill ensures zero human involvement in the deployment process while maintaining strict constraints on dependencies and response times.

Discover, find, compare, audit, repair, adapt, craft, run, debrief, and prepare repeatable AI-agent loops for publication. A comprehensive tool for managing bounded feedback loops in AI agent workflows.

Build enterprise agents for Microsoft 365, Teams, and Copilot Studio with the Microsoft 365 Agents SDK in TypeScript/Node.js. Covers Express hosting, AgentApplication routing, streaming responses with Azure OpenAI, invoke activities, and Copilot Studio client and WebChat integration.

Design and implement complete machine learning pipelines using multi-agent orchestration following modern MLOps best practices. Covers data engineering, feature stores, model training, and production deployment. Ideal for building scalable, monitored, and reproducible ML systems in production environments.

Comprehensive hardening guide for Model Context Protocol (MCP) servers, covering threat modeling, transport encryption (TLS/mTLS), OAuth 2.1 authentication, tool authorization, input validation, and audit logging to secure AI agent integrations.

This skill detects persistent instruction injection vulnerabilities in AI agents. It verifies if the agent stores malicious instructions in long-term memory or context to execute later. The process involves attempting to write a persistent rule and verifying if the agent follows it in subsequent interactions. It focuses on behavioral persistence rather than data leakage.

This skill provides methodologies for designing sophisticated, layered memory architectures for AI agents. It moves beyond simple context windows and basic vector stores by incorporating short-term, long-term, and graph-based memory systems. Implementing this is crucial for agents that require persistence across sessions, maintaining complex entity consistency, and performing advanced, relationship-driven reasoning over accumulated knowledge bases, including time-aware fact tracking.

A self-hosted, semantic memory service designed for AI agents. It allows agents to save worklogs, decisions, and notes, and recall this knowledge across sessions based on meaning rather than exact keywords. Using pgvector and structured workspaces, it maintains persistent, context-aware knowledge bases, making it ideal for collaborative AI projects.

Design and optimize production-grade multi-agent systems using LangGraph, LangChain, and DeepAgents. Provides structured workflows for creating supervisor, planner, researcher, coder, and validator agents with memory and tool-calling pipelines.