Javascript agents

harness-neutral Constructive database tools — provisioning, schema, policies, records, templates and codegen as HarnessTools any adapter can bind

Provider-agnostic LLM adapter with streaming and multi-turn support

The append-only agent run log — pi session entries stored verbatim under a run/seq wrapper, with projections for transcript, usage, tool state and resumable sessions

The runtime layer for Agentic AI. Persistent memory, autonomous agents, and workflows that survive every session.

Framework for creating agentic ibGib web apps. Contains plumbing for ibgib components, agentic framework (currently only Gemini implemented), web-based IndexedDB storage substrate, and more.

A Pi extension that keeps long-running agentic sessions on track with background memory, contemplation, and structural review.

Agentic SDLC workflow toolkit — installs role-specific AI agents + CodeGraph knowledge base into any workspace.

Agentic Installer: self-contained CLI that installs AI-agent guidance (rules, skills, workflows, prompts) into any project. This npm package is a thin launcher that downloads the prebuilt Rust binary for your platform from GitHub Releases.

Installs and maintains a repository's agentic development system, composed from modules.

HeartFlow (心虫) discriminator tool for CrewAI — gate agent output before it reaches users.

Grounded Intelligence for AI agents. MCP server with real-time web search, evidence extraction, verification, and structured citations. The research infrastructure for LangChain, CrewAI, and custom agents.

Reserved for LastSearch — verified search infrastructure for AI agents. Full release imminent. https://lastsearch.ai

Intertrace runtime protection SDK — authorize+permit tool gates and agent-builder adapters

Measure how much of an AI coding agent's output is still in a git repo, and compare it to the humans on the same repo.

Trust Agent Protocol (TAP) SDK — agent registration, passport, policy evaluation, human review, audit, and evidence

AG-UI protocol client for Angular — SSE streaming, signals store, client-side tools, and human-in-the-loop approvals. Works with any AG-UI compatible agent backend.

BTP v3.0 The AI Agent Execution Gateway — Sub-35µs In-Process Tool Gating, AST Invariants, Secret Scrubbing, Loop Damping & Offline Merkle Receipts

Deterministic runtime evidence and diagnostics for LangGraph TypeScript workflows.

Provider-neutral graph primitives built on LangGraph and LangChain.

Context-aware LLM agents and pipelines over the LangGraph functional API, with pluggable memory and storage-independent persistence.

The DevPilot conductor as a LangGraph agent — plan, review, dispatch and advance a wave plan as one resumable graph

LangChain and LangGraph tools for txn.dev. Agent payments and hiring humans.

udctl — the AI agent CLI: let Claude Code, Cursor & Codex manage tasks, notes and expenses from the terminal

TapTap AI Agent CLI

Lapis Notes provider-agnostic AI agent plugin

The AI Agent Firewall. Enforce signed authority before consequential actions and protect sensitive data at the execution boundary. Free API key required.

Standalone cross-provider translation router for AI-agent hooks with ordered fallbacks and shared caches

Token-efficient coverage report compressor for LLM/agent consumption. Reads lcov, outputs compact TSV.

A pixel-sprite editor whose entire feature set is reachable from JavaScript, so an LLM agent can draw, animate and export sprites without ever touching the UI.

Veryfront first-party extension package for ext-llm-anthropic

Seamless API-key rotation for DeepSeek Harness LLM providers when subscription limits are hit

WASM Node.js bindings for libdatadog

Provider-agnostic LangChain and LangGraph tools, finite nodes, and an analyst for @xbbg/xnews.

Autonomous API Schema & Reliability Inspector CLI. Probes OpenAPI specs, generates test suites, executes real-time HTTP calls, and exports Postman collections.

FrontTerrain — The Onboarding Co-Pilot for New Codebases

本地优先的多 Agent 协作与自动开发工作区，集成任务编排、记忆上下文、MCP/Skill、TestAgent、Git、飞书、知识库和终端

Many agents. One runtime. Separate lives. Run distinct local AI agents — each with its own soul, memory, secrets, skills, workspace, and autonomy level.

Pinned control plane: directory, placement, and HTTP proxy

Shared transport contracts for AI App Assistant packages

Lightning-fast, zero-lag AI coding assistant for TypeScript with instant undo and drag-and-drop customization

TypeScript SDK for programmatic control of GitHub Copilot CLI via JSON-RPC

Local multi-agent orchestration framework with an interactive pixel-art tech startup office visual dashboard

Zero-runtime multi-agent orchestration convention for Claude Code, OpenCode, Codex, Antigravity, and Cursor

Evidence-driven multi-agent routing with deterministic safety gates.

The open-source compiler and multi-agent runtime for coding agents

TokenKoding Code — agentic coding CLI. Talks only to api.tokenkoding.id.

The standard for agentic identity on Nostr, plus a thin, dependency-light conformance primitive: the identity types, the act-outward gate predicate (owned + self + verified + funded), and a pure structural conformance checker. ESM + CommonJS.

SquadOS — Operating System for AI Agentic Software Company (MCP Server)

Verification-first agentic bug fixing system

MCP server for the numkit numerical computing engine — MATLAB-compatible code execution for AI agents with a persistent workspace.

Express + Zod router that validates requests, and infers OpenAPI response schemas directly from your sendSuccess()/sendError() call sites — no separate response schema to write or keep in sync.

YÖK Akademik — Canlı Akademik Veri Kazıma, AI Agent (MCP) ve Dışa Aktarım Paketi

The file-format table the add-track guessers and the CLI both read — no framework deps

DeepSeek Harness QQ Bot Adapter - C2C Single Chat, Session Sync, Interactive Action Board & Custom Menu

Official AI Agent and LLM Function Calling Tools for BaZi API (OpenAI, LangChain, Vercel AI SDK)

APort Repository Guard and pre-action authorization guardrails for GitHub, Claude Code, Cursor, OpenClaw, LangChain, CrewAI, DeerFlow, and n8n

Weaver — Coding agent dệt plan, code và test thành phần mềm chạy được. Xây dựng với LangGraph: Planner → Analyzer → Coder → Tester → Reviewer, tự sửa khi test fail.

Detect if code is running in an AI agent or automated development environment

BitFun CLI - terminal AI agent

AI Extensions: a package-manager-style CLI for project-local AI-agent extensions.

Local deterministic pre-attachment security scanner for MCP and AI-agent tool manifests.

AgenticPool.net — Decentralized AI agent favor exchange network powered by Duckies & A2A Trust Protocol

Bring your own AI agent into Google Meet, Zoom & Discord voice channels, as a real voice participant.

Terminal AI Agent with desktop pet Arona — eye-tracking pupils, voice cloning, Computer Use, TTS/STT, MCP.

Reusable AI agent chat server library for Hono + Bun

Haven MCP server for local, non-custodial AI-agent payments — pay x402 (HTTP 402) APIs within on-chain-enforced budgets.

Cadet Brainstem — an orchestration + measurement layer that reduces AI-agent context consumption.

Word-faithful React DOCX editor and viewer — full OOXML fidelity, real-time human + AI agent collaboration.

Let Agents Chat — MCP server for AI agent communication

Local cockpit for running and tracking AI agent tasks in your repo. Uses your logged-in `claude` CLI and `gh` — zero config, zero database.

Reusable AI agent abstraction built on Google ADK

LCDD MCP Server — AI agent integration via Model Context Protocol

GemDesign CLI - AI agent 驱动的原型设计命令行工具

Tiny provider-neutral authorization checkpoint for AI-agent actions: allow / deny / require_approval.

A fast CLI tool to sync AI agent configurations and MCP servers across Claude, Copilot, Cursor, and more using symbolic links.

Swarm DAO core — pure business logic for AI agent governance

Snyk security scans for SAST, SCA and Secrets, with an AI agent-optimized interface

Model Context Protocol (MCP) Server and AI Agent Skill Suite for OpenRP.ai

Find what is wrong with the AI agent configuration your repository already has

ArcFusion AI agent UI components

MCP server for the 3x-ui Xray/V2Ray proxy panel: manage inbounds, clients, routing, nodes and Xray itself from an LLM agent.

Deterministic semantic record and replay for LLM agent tests

Active Loader-backed plugin package inventory for official DeepSeek LLM API requests

General-purpose agent with transport abstraction, state management, and attachment support

Abstract compaction service seam (ctx.compaction) for the DeepSeek Harness

Local bridge for TeamShare agents - CLI session runner, auto-wake WebSocket daemon, and teamshare:// protocol handler.

Agent runtime for faapi — LLM-driven ReAct loop over scanned agent / tool registries

An ESLint plugin that tries its best to prevent your coding agents from pooping your bed.

Google-OAuth-gated LLM gateway: verify a Google ID token, then run a Gemini (Vertex AI) completion for the verified caller.

智能体：自动将代码项目国际化的扫描-分类-提取-翻译-替换-验证管线（不限于 ERP，适配 Vue/React/TS/JS/HTML）

Anchor your LangChain.js agent's run tree as it executes: one callback handler turns every chain, LLM and tool step into a signed evidence envelope, sent to Sanning and witnessed on an interval to a permanent public record. Prompts and outputs are hashed

OpenAI provider tokens and model factory wiring for Nest LangChain.

Drop-in replacement for ChatGoogleGenerativeAI that automatically fixes schema compatibility issues with MCP tools and Gemini API

Reusable model provider adapters and streaming runtime from OpenClaw

Ultra-fast local agent shell for Windows.

ChatGPT OAuth and Codex models for DeepSeek Harness.

Canonical Pi agent loop for Assembly Line behind the internal AgentHarness runtime contract.

OpenAI toolkit for Pi: Codex Remote Context windows, remote compaction v2, routed Web Search, image gen, auto mode.

一个纯净、极简、易于使用的 Agent 框架

Unified LLM API with automatic model discovery and provider configuration