name: libagent description: > libagent - Agent orchestration library for conversational AI. AgentMind class coordinates LLM completions, memory management, tool execution, and multi-turn conversations. AgentAction handles tool calls and action processing. Use for building chat agents, RAG pipelines, and AI assistants. Integrates with libmemory, librpc, and libllm

libagent Skill

When to Use

Building conversational AI agents with tool capabilities
Coordinating LLM completions with memory context
Processing multi-turn conversations with state
Implementing RAG (retrieval-augmented generation) pipelines

Key Concepts

AgentMind: Core reasoning engine that processes requests through LLM completions, manages conversation state, and coordinates tool execution.

AgentAction: Handles individual tool calls, executes actions, and returns results to the conversation flow.

Usage Patterns

Pattern 1: Basic agent request

import { AgentMind } from "@copilot-ld/libagent";

const mind = new AgentMind(memoryClient, llmClient, toolClient);
const response = await mind.process({
  resourceId: conversationId,
  content: "What is the weather?",
});

Pattern 2: Streaming responses

for await (const chunk of mind.stream(request)) {
  process.stdout.write(chunk.content);
}

Integration

Works with

libmemory

for context windows,

librpc

for gRPC clients, and

libllm

for completions. Used by the Agent service.

name: libagent description: > libagent - Agent orchestration library for conversational AI. AgentMind class coordinates LLM completions, memory management, tool execution, and multi-turn conversations. AgentAction handles tool calls and action processing. Use for building chat agents, RAG pipelines, and AI assistants. Integrates with libmemory, librpc, and libllm

libagent Skill

When to Use

Building conversational AI agents with tool capabilities
Coordinating LLM completions with memory context
Processing multi-turn conversations with state
Implementing RAG (retrieval-augmented generation) pipelines

Key Concepts

AgentMind: Core reasoning engine that processes requests through LLM completions, manages conversation state, and coordinates tool execution.

AgentAction: Handles individual tool calls, executes actions, and returns results to the conversation flow.

Usage Patterns

Pattern 1: Basic agent request

import { AgentMind } from "@copilot-ld/libagent";

const mind = new AgentMind(memoryClient, llmClient, toolClient);
const response = await mind.process({
  resourceId: conversationId,
  content: "What is the weather?",
});

Pattern 2: Streaming responses

for await (const chunk of mind.stream(request)) {
  process.stdout.write(chunk.content);
}

Integration

Works with

libmemory

for context windows,

librpc

for gRPC clients, and

libllm

for completions. Used by the Agent service.

libagent

libagent Skill

When to Use

Key Concepts

Usage Patterns

Pattern 1: Basic agent request

Pattern 2: Streaming responses

Integration

Related Skills

Markdown Converter

Nano Banana Pro

1password

libagent Skill

When to Use

Key Concepts

Usage Patterns

Pattern 1: Basic agent request

Pattern 2: Streaming responses

Integration