agentic-skills

Slash commands for Claude Code. Each `.md` file in this directory becomes a `/command-name` you can invoke directly.

Apply the Adaptation pattern — dynamically modify agent behavior based on feedback or performance metrics

Apply the Inter-Agent Communication pattern — enable agents to exchange messages and collaborate

Apply the Prompt Chaining pattern — chain LLM calls where each output feeds the next

Apply the Evaluation & Monitoring pattern — measure agent performance, reliability, and cost

Apply the Exploration & Discovery pattern — autonomously research, hypothesize, and explore solution spaces

Apply the Guardrails & Safety pattern — inspect inputs/outputs to prevent misuse and harmful content

Apply the Exception Handling & Recovery pattern — detect failures and execute fallback logic

Apply the Human-in-the-Loop pattern — pause execution for human approval or input on critical actions

Apply the Model Context Protocol (MCP) pattern — connect AI to external data and tools securely

Apply the Memory Management pattern — persist and retrieve state across agent interactions

Apply the Multi-Agent Collaboration pattern — coordinate specialized agents on complex problems

Apply the Resource-Aware Optimization pattern — select the most efficient model or tool per task

Apply the Parallelization pattern — execute multiple agent tasks concurrently for speed and diversity

Apply the Planning pattern — formulate structured action sequences before execution

Apply the Prioritization pattern — assess task urgency and importance to organize execution

Apply the Knowledge Retrieval (RAG) pattern — augment generation with external knowledge

Apply Reasoning Techniques — use Chain of Thought to improve performance on complex tasks

Apply the Reflection pattern — evaluate and critique agent output to iteratively improve quality

Apply the Routing pattern — classify inputs and direct them to the right specialized agent

Apply the Goal Setting & Monitoring pattern — define success criteria and track progress iteratively

Apply the Tool Use pattern — enable LLMs to execute actions and retrieve information from external systems

No agents are defined for this plugin yet. Agent definitions and orchestration logic can be added here.

A dynamic pattern where an agent system modifies its own behavior, prompts, or tools over time based on feedback or performance metrics. Use when user asks to "make my agent adaptive", "add learning capabilities", "self-improving agent", or mentions adaptive behavior, online learning, or feedback loops.

Systems for quantitatively and qualitatively measuring agent performance, reliability, and cost. Use when user asks to "evaluate agent performance", "benchmark my agent", "test agent quality", or mentions agent metrics, scoring, or performance assessment.

Patterns for ensuring system resilience by detecting failures (API errors, hallucinations, validation errors) and executing predefined fallback logic. Use when user asks to "handle agent errors", "add error recovery", "make my agent fault-tolerant", or mentions exception handling, graceful degradation, or retry logic.

An open-ended pattern where agents autonomously conduct research, generate hypotheses, and explore solution spaces without a predefined path. Use when user asks to "add exploration to my agent", "balance exploration and exploitation", or mentions curiosity-driven, search strategies, or novelty seeking.

An iterative pattern where an agent defines clear success criteria and continuously evaluates its progress, adjusting its actions until the goal is achieved. Use when user asks to "set agent goals", "define objectives for my agent", "goal decomposition", or mentions goal hierarchies, sub-goals, or objective functions.

A defensive pattern where inputs and outputs are inspected by dedicated safety agents or rules to preventing malicious use, jailbreaks, and harmful content. Use when user asks to "add safety checks", "set up guardrails", "prevent harmful outputs", or mentions agent boundaries, output validation, or content filtering.

A hybrid pattern where the system pauses execution to request human approval, input, or disambiguation before proceeding with critical actions. Use when user asks to "add human approval", "require human review", "human-in-the-loop", or mentions approval workflows, human oversight, or escalation.

Protocols and patterns that allow independent agents to exchange messages, negotiate, and collaborate across network boundaries or process isolation. Use when user asks to "make agents communicate", "agent messaging", "inter-agent protocol", or mentions agent coordination, message passing, or shared state.

An interoperability standard that allows AI models to connect to external data and tools securely and consistently, decoupling the tool implementation from the agent. Use when user asks about "MCP servers", "model context protocol", "connect tools to my agent", or mentions MCP integration, tool servers, or context protocol.

Techniques for persisting, retrieving, and managing state across agent interactions, enabling long-coherency and personalization. Use when user asks to "add memory to my agent", "persistent context", "conversation history", or mentions long-term memory, memory retrieval, or context windows.

A structural pattern where multiple specialized agents communicate and coordinate to solve a problem that is too complex for a single agent. Use when user asks to "build a multi-agent system", "agents working together", "agent collaboration", or mentions team of agents, distributed agents, or swarm.

A concurrency pattern where multiple agent tasks are executed at the same time to speed up processing or gather diverse perspectives. Use when user asks to "run agents in parallel", "parallelize tasks", "concurrent execution", or mentions parallel processing, fan-out, or batch execution.

A high-level cognitive pattern where an agent formulates a structured sequence of actions (a plan) before executing any of them, ensuring goal-directed behavior. Use when user asks to "add planning to my agent", "task planning", "agent planning", or mentions plan generation, plan execution, or step-by-step planning.

A management pattern where an agent assesses the urgency and importance of incoming tasks to organize a dynamic execution queue. Use when user asks to "prioritize tasks", "rank agent actions", "task ordering", or mentions priority queues, urgency scoring, or triage.

A fundamental pattern where the output of one LLM call serves as the input for the next, enabling complex tasks to be broken down into manageable sequential steps. Use when user asks to "chain prompts together", "multi-step prompts", "prompt pipeline", or mentions sequential prompts, prompt workflows, or chain-of-thought.

A pattern that augments the model's generation by retrieving relevant documents from an external knowledge base, ensuring factual accuracy and access to private data. Use when user asks to "add RAG to my agent", "retrieval augmented generation", "search and answer", or mentions document retrieval, knowledge bases, or semantic search.

Prompting patterns that encourage the model to articulate its specific thought process (Chain of Thought) to improve performance on complex logical, mathematical, or reasoning tasks. Use when user asks to "improve agent reasoning", "add chain-of-thought", "logical reasoning", or mentions inference, deductive reasoning, or structured thinking.

A recursive pattern where an agent evaluates and critiques its own output to iteratively improve quality and catch errors. Use when user asks to "add self-reflection", "agent introspection", "self-critique", or mentions self-evaluation, meta-cognition, or quality self-assessment.

A routing pattern that optimizes for cost and latency by dynamically selecting the most efficient model or tool for a given task complexity. Use when user asks to "optimize agent resources", "reduce token usage", "efficient agent", or mentions cost optimization, latency reduction, or throughput.

A control flow pattern where a central component classifies an input request and directs it to the most appropriate specialized agent or tool. Use when user asks to "route between agents", "agent routing", "task dispatch", or mentions classifier routing, intent detection, or agent selection.

The capability that transforms an LLM from a text generator into an agent by allowing it to execute actions and retrieve information from the real world. Use when user asks to "add tools to my agent", "function calling", "tool integration", or mentions API calls, external tools, or agent capabilities.