Agents Core contains the database schema, types, and validation schemas for Inkeep Agent Framework, along with core components.
The OpenAI Agents SDK is a lightweight yet powerful framework for building multi-agent workflows.
Headless TypeScript framework for building durable, deployable autonomous agents — core SDK.
The app framework built for AI coding agents - core runtime
Autonomys Agents Core
Make your Saleor store shoppable by AI agents. Core protocol types, formatters, and payment handler interface for UCP and ACP.
Scaffold a Bun agent app powered by @lucid-agents/core.
Core services for Reactive Agents — EventBus, AgentService, TaskService, and shared types
Core data models and business logic for Epic Flow AI Agents
The First UI Library Built for AI Coding Agents - Core CSS and JavaScript
MCP Server para o core protegido do Claude Agents Ecosystem v1.4.6 — 54 agentes especializados. v1.4.6 RELEASE MAJOR: 13 issues consolidadas (255-267) — Spawn fixes + Pacote A Modularidade (file-size-guard, executor-file-scope, agents.md per-subdir, templ
The OpenAI Agents SDK is a lightweight yet powerful framework for building multi-agent workflows.
Core types, schemas, and utilities for some-useful-agents
Build typed agent HTTP apps
ENS-native self-evolving agent framework with persistent 0G tool memory
LangGraph
Camoufox-backed browser automation CLI for AI agents (core agent-browser compatible)
No description provided.
Shared primitives for @ar-agents/* packages: typed error base, tool middleware (compose, withMetrics, withRetry, withTimeout, withApproval HITL gate), telemetry hooks, adapter contract conventions. Zero runtime deps beyond Vercel AI SDK's tool shape.
A JavaScript framework for orchestrating multiple AI-driven agents to handle complex tasks like data processing, code generation, chart creation, and image generation
A TUI-first coding agent CLI built on top of agents-core
Hud's Node SDK
Agent Kernel (formerly KavachOS) — seccomp-bpf + Falco kernel enforcement for AI agents. Core kernel-enforcement primitive of the xShieldAI Posture Suite.
Maps proxy protocols to `http.Agent` implementations
Core traits, data models, and prompt primitives for building deep agents.
AWS integrations for the Rust deep agents SDK.
Async runtime orchestration for Rust deep agents.
High-performance Rust SDK for composing reusable AI agents with custom tools, sub-agents, and prompts
Reusable tools and utilities for Rust deep agents.
Core types and traits for AI Agents framework
Headless engine for ascii-agents — state, sprites, layout
Core types for the atomr-agents framework: ids, budgets, context, events, errors.
Agent Framework for Building Autonomous Agents
Core routing, policy, and orchestration primitives for codex-agents
Core runtime primitives for openai-agents-rust.
Core SimpleAgents client API integrating providers and healing
Parses the typically messy content in Darwin Core terms that contain people names
Ruby-native LLM agent framework with provider adapters (Anthropic, OpenAI, Google), tool calling, streaming, and session persistence.
Clyro - The intelligent core for AI agents
pikuri-core is the lean, audit-friendly foundation of the pikuri family: Pikuri::Agent (a thin wrapper around ruby_llm's chat loop) with its Configurator + Extension protocol, the strict Pikuri::Tool framework, a listener surface for rendering / budgets / sub-agents, and four bundled stateless tools (calculator, web search, web scrape, fetch). Extensions (skills, MCP, workspace, coding stack, named-agent personas) live in sibling gems so a privacy-conscious user can install just this core and audit a minimal dependency tree. For the convenience bundle that pulls in everything, see the +pikuri+ metagem.
OpenClacky is a 100% open-source (MIT) AI Agent that matches Claude Code on capability at comparable cost. Built on two years of agentic R&D, it ships with a minimal 16-tool core, near-100% prompt cache hit rate, idle-time auto-compression, and BYOK support for any OpenAI-compatible model. A natural-language Skill system lets you create, evolve, and monetize agent capabilities without writing code, plus first-class integrations with Feishu, WeCom, WeChat, Discord, and Telegram.
pikuri-skills implements the Agent Skills standard (agentskills.io) for pikuri-core agents: a +Pikuri::Skill::Catalog+ that discovers skill folders under .pikuri/skills, .claude/skills, and .agents/skills; a +Pikuri::Skill::SkillTool+ that loads a skill's body on demand; and a +Pikuri::Skill::Extension+ that wires both into a +Pikuri::Agent+ via +c.add_extension(...)+ in the +Agent.new+ block.
Devex provides a unified `dx` command for common development tasks. Features include: - CLI framework with automatic help generation and nested subcommands - Agent-aware output (detects AI agents and adapts output format) - Environment orchestration (mise, bundle exec, dotenv integration) - Project path conventions with fail-fast feedback - Zero-dependency support library (Path class, ANSI colors, core extensions)
If you're building a game, you need your game agents and characters to move on their own. A standard way of doing this is with 'steering behaviors'. The seminal paper by Craig Reynolds established a core set of steering behaviors that could be utilized for a variety of common movement tasks and natural behaviors. This Ruby library can accomplish many/most of those tasks for your Ruby / JRuby game. The basic behaviors can be layered for more complicated and advanced behaviors, such as flocking and crowd movement.
pikuri-tasks gives a pikuri-core agent an in-memory task list it can use to plan and track multi-step work. A +Pikuri::Tasks::List+ holds the per-Agent state; four tools (+task_create+, +task_in_progress+, +task_completed+, +task_delete+) mutate it via content-as-identifier (no item IDs to hallucinate). +Pikuri::Tasks::Extension+ wires the list and tools onto an +Pikuri::Agent+ via +c.add_extension(...)+ inside the +Agent.new+ block. The list lives in process memory only — nothing is written to disk. Sub-agents do not inherit the parent's list (consistent with the "sub-agents do not inherit extensions" rule).
Engram gives AI agents durable, long-term memory. It recalls relevant facts about a user and injects them into the prompt, so an agent appears to remember across sessions. Framework-agnostic core with a ports-and-adapters design; first-class Rails and RubyLLM integration. Your memories live in your own database — no external memory service.
pikuri-mcp adds Model Context Protocol support to pikuri-core agents: a +Pikuri::Mcp::Registry+ for declaring stdio + HTTP MCP servers, the +Pikuri::Mcp::Servers+ runtime that spawns them, a +Pikuri::Mcp::Synthesizer+ that LLM-fills missing server descriptions, a +Pikuri::Mcp::Verifier+ that screens server surfaces for prompt-injection patterns before any tool is advertised to the LLM, and a +Pikuri::Mcp::Extension+ that wires everything into a +Pikuri::Agent+ via +c.add_extension(...)+ in the +Agent.new+ block.
pikuri-assistant ships the demo +bin/pikuri-assistant+ binary on top of pikuri-core + pikuri-skills + pikuri-mcp: an interactive chat agent the user hand-edits to declare a +Pikuri::Mcp::Registry+ of MCP servers. The +Pikuri.prompt+ search path picks up this gem's +prompts/pikuri-assistant.txt+ automatically on require — same multi-gem prompt-dir pattern as pikuri-code.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.
No description provided.