context-first

Use when validating or diagnosing the cortex environment (consistency, duplicates, redundancy, optimization) - provides a diagnostic workflow and reporting structure for context health checks.

Always Auto-invoked skill that creates/updates workspace AGENTS.md to instruct the agent to always search for existing skills before attempting any scientific task.

context-monitor

Parallel historical context gathering using G-4 CONTEXT. Deploy 6 probes to gather all relevant context before complex work.

MCMC convergence diagnostics using CmdStanPy and ArviZ

Integrates human agents into automated workflows for human-in-the-loop interactions. Use this when humans need to respond alongside AI agents, handle escalations, or provide context that AI cannot determine.

Deploy to Coolify with best practices

Invoke COORD_OPS for releases, documentation, and CI/CD operations

Automatically activated when user mentions multi-project coordination, cross-project dependencies, portfolio management, roadmap planning, resource allocation across projects, or asks to coordinate/manage multiple projects simultaneously. Provides strategic project coordination expertise.

This skill intelligently assigns GitHub issues to the most appropriate GitHub Copilot agent(s) based on issue type, labels, technical domain, and dependencies.

This skill manages complex workflows involving multiple GitHub Copilot agents working on related issues. It handles sequencing, handoffs, dependency resolution, and progress tracking for coordinated m

**Goal:** Determine how to proceed when major change is discovered

Self-corrects skill and command instructions after resolving errors. Use when Claude errors during skill/command execution and finds a solution, or when user indicates a mistake was made. Finds the relevant instruction file and applies succinct fixes.

Operate the Agent Hive Cortex orchestration engine. Use this skill when running Cortex commands, analyzing project dependencies, finding ready work, understanding the orchestration system, or troubleshooting Cortex issues.

計算交易成本對風險報酬比的非線性衰減影響。將固定佣金與點差整合為「成本密度」指標，揭示停損大小與策略效率的雙曲線關係，識別「獲利事件視界」閾值。

Techniques for gaining visibility into cloud spending, attributing costs to business units, and detecting financial anomalies.

The First Sovereign AI Agent Democracy & Judiciary. A protocol for autonomous agent governance, dispute resolution, and legislative deliberation.

create-adr-spike

<instructions>

**Goal:** Create comprehensive architecture decisions through collaborative

Creates a new custom slash command file. Use when: user wants to create a new command, add a slash command, or define a custom command.

**Goal:** Break down architecture into implementable epics with user stories.

Create properly named feature branch from development with remote tracking, following WescoBar naming conventions and git best practices

>

|-

**Goal:** Create comprehensive product briefs through collaborative step-by-step

Use to decompose user stories into individual development tasks. Creates task documents that can be assigned, estimated, and tracked.

Creates a sub-issue linked to a parent GitHub issue. Use when: user wants to create a sub-issue, break down an issue, or add child task.

This skill should be used when the user asks to 'create a task from', 'add issue for', 'track work on', or provides conversational task descriptions like 'I need to fix the blog bug' or 'remind me to update the docs'. Converts natural language task descriptions into structured beads with automatic categorization, priority inference, and optional blocking relationships.

Use when setting up a new open-ended goal for autonomy tracking, before starting the first iteration

Execute feature branch creation with intelligent naming, automatic type detection, and sequential numbering.

Every GitHub issue should be well-researched and clearly documented. This skill ensures issues contain enough context for anyone to understand and implement them.

Create GitHub Issue knowledge entries. TRIGGERS - share knowledge, document solution, create how-to, add to knowledge base.

Генерация рекламных изображений через Gemini. Используй для создания креативов с текстом.

Professional skepticism with pattern enforcement. Verify before agreeing, challenge violations, propose instead of asking. Concise output, research before asking, answer questions literally.

You are a crontab manager for this dotfiles repository. Help manage scheduled tasks in `config/crontab` with precision and care.

處理跨 Bounded Context 的依賴關係。當需求涉及多個限界上下文（如權限管理、付款、通知）時觸發，設計 Anti-Corruption Layer 和 Context Mapping。

Structured methods for finding connections across disciplines. Use when exploring how concepts from one field illuminate another, seeking novel applications, or analyzing structural similarities between domains.

Coordinate changes across project-beta repositories when updating runner configurations. Ensures workflow labels match runner scale set names. Use when changing runnerScaleSetName or deploying new runner pools.

Use when designing or implementing cross-service communication, data synchronization, or service boundary patterns.

Synchronizes Cursor Rules (.mdc files in .cursor/rules/) to CLAUDE.md by generating a Rules section with context-efficient descriptions and usage instructions. Use when setting up Cursor Rules for the first time, after adding or modifying rules, or when the Rules section in CLAUDE.md is missing or outdated.

Customer-facing documentation

Эксперт Customer Success. Используй для onboarding, retention, upsell и customer health scoring.

Эксперт customer training. Используй для onboarding программ, training materials и certification paths.

Google ADK (Agent Development Kit) architecture and core concepts. Covers agent types (LLM, Sequential, Parallel, Loop, Multi-agent), tool ecosystem, execution flows, deployment options, and observability. Model-agnostic framework for production AI agents. Use when understanding ADK architecture, selecting agent patterns, designing agent systems, or deploying to Vertex AI.

cybersecurity-analyst

AI深度分析语音记录，揭示行为模式、认知偏见和战略盲点。不只是数据汇总，而是直击本质的战略洞察和否定性指导。提取可沉淀的核心原则和工作偏好，构建个人工作模式知识库。

Install, configure, and manage the Claude Code Damage Control security hooks system. Use when user mentions damage control, security hooks, protected paths, blocked commands, install security, or modify protection settings.