AI Engineering

ArcKit Build orchestrates parallel ArcKit artefact generation using subagent isolation, computing dependency DAGs and committing waves. Use it to bulk-build or resume ArcKit projects instead of running individual /arckit:* commands.

Auto Memory creates and maintains a structured persistent memory system for Claude Code projects. Use it when asked to set up MEMORY.md, topic files, or improve cross-session knowledge retention.

PlantUML Syntax provides reference material for C4-PlantUML, sequence, class, activity, state, ER, component, deployment, and use case diagrams. Use it to answer syntax questions, troubleshoot rendering errors, and apply styling or layout constraints.

Brainstorm is a unified brainstorming skill that guides Claude Code through interactive framework generation, parallel multi-role analysis, and cross-role synthesis. Use it when you need structured ideation, role-based analysis, or feature specifications for a project.

Delegation Check validates workflow delegation prompts against agent role definitions to detect content separation violations. Use it to catch conflicts, duplication, boundary leaks, and missing contracts in Claude Code skills.

Memory Capture routes memory capture requests to either a full session compaction or quick tip note-taking, saving structured output via core memory import for later recovery. Use it when you need to save session context or capture quick notes during agent workflows.

Memory Manage routes project memory (CLAUDE.md) updates and documentation (API.md/README.md) generation to specialized subcommands based on user input or interactive prompts. Use it to update full, related, or single modules, and to generate full or incremental docs.

Prompt Generator generates or converts Claude Code prompt files, including command orchestrators, skill files, and agent role definitions. Use it when you need to create or restyle Claude Code commands, skills, or agents with consistent structure.

Skill Generator creates new Claude Code skills from user requirements, supporting sequential and autonomous execution modes for phase-based workflows. Use it to scaffold skill directories, generate phase files, and validate outputs according to design specs.

Skill Simplify analyzes SKILL.md files for redundancy and optimizes their content while verifying that no functionality is lost.

Skill Tuning diagnoses and fixes skill execution issues such as context explosion, long-tail forgetting, data flow disruption, and agent coordination failures. Use it to run diagnosis, apply fixes, and verify quality gates for any skill.

Team Architecture Optimization orchestrates a multi-agent pipeline to analyze a codebase, design a refactoring plan, implement changes, validate improvements, and review code quality. Use it when you need coordinated architecture refactoring across specialized workers.