AI Agents

Acts as an autonomous AI teammate for business teams, carrying out complete marketing, sales, engineering, and operations tasks rather than serving only as an assistant.

Spring-based personal AI operating system where a single agent serves web admin, desktop app, embeddable widget, and eight IM channels from one JAR deployment.

Lightweight terminal coding assistant that replicates Claude Code-style workflows with tool loops and an extensible TUI architecture. Provides TypeScript, Python, and Rust implementations to help learners study and experiment with agent design.

Swift-native macOS coding agent built for DeepSeek, paired with a cross-platform command-line tool that runs autonomously to deliver working code.

Generalist autonomous research agent that runs experiments, surveys literature, and iteratively optimizes hypotheses. A hypothesis tree preserves validated gains for researchers pursuing open-ended scientific exploration.

High-fidelity vulnerability scanner combining agentic AI with fast native execution, offering deterministic scans and autonomous planned exploitation for code auditing and security review.

An operations AI agent that connects to infrastructure, investigates alerts to find root causes, and executes fixes directly from Slack, Telegram, Lark, or DingTalk chat interfaces.

An open-source autonomous Android agent that plans multi-step tasks and operates your phone apps, messages, and settings for you.

On-device AI assistant that autonomously operates Android phones using local models without cloud services or API keys. Maintains task state across applications and chat workflows to help users automate everyday mobile interactions privately.

An open Claude Code-style coding agent with persistent memory, desktop UI, remote control, and computer use.

Agentic AI security tool applying proactive, attacker-first analysis directly to source code, validating exploit paths and generating evidence-backed findings with remediation guidance for security engineers.

Open-source template for building file-system-backed knowledge agents that stay synchronized without vector databases. Helps developers create deployable chat and bot applications grounded in continuously updated organizational documents.

Coordinates a fleet of AI software-engineering agents that plan, write code, test, and ship production-grade pull requests autonomously within the AgentField development environment.

Decoupled multi-agent framework for text-to-CAD generation, producing editable parts and multi-part assemblies with validation and export for simulation, inspection, and manufacturing workflows.

Human-agent collaborative workspace system for recruiting and orchestrating multiple agents on shared team tasks, with role permissions, approval workflows, and audit controls.

Acts as a practical AI penetration-testing engineer for authorized assessments, combining cloud-detection modules and deterministic execution across major cloud providers built on the Claude Agent SDK.

Terminal-resident autonomous AI agent that builds apps and backends, manages files, performs deep research, and schedules tasks for developers working in TUI, CLI, or daemon mode.

Enterprise multi-agent IT operations platform with LLM-powered auto-remediation for Zabbix and Prometheus alerts, deployable via Docker for operations teams managing large infrastructures.

Open-source, self-hosted platform for governed AI agent teams that autonomously handle routine business operations. Emphasizes verifiable outputs, policy controls, and secure sandboxing to help organizations automate workflows with trustworthy automation.

Full-featured terminal programming-agent runtime offering independent judgment, cognitive safeguards, decaying memory, and prefix-cache optimizations for long DeepSeek sessions and developer partnership workflows.

Provides a minimal self-hosted coding-agent command-line interface that reads and writes files and executes shell commands to complete programming tasks for individual developers.

Self-hosted AI companion robot with long-term memory, voice interaction, camera vision, and smart-home integration, controlled through a mobile app for personal home use.

An AI-native A-share research workstation that auto-collects market views and delivers evidence-backed stock and portfolio analysis.

Mobile autonomous penetration-testing agent runs local AI models directly on phones to discover host vulnerabilities and generate test reports for security practitioners and learners.

Emerging human-behavior simulation capability in LLM agents

An autonomous agent on Solana for managing Meteora liquidity, continuously screening pools, deploying capital, and automatically rebalancing or closing positions based on real-time yield.

Terminal-based multi-agent autonomous penetration-testing assistant that runs reconnaissance, scanning, and exploitation in parallel while maintaining state for security professionals testing workflows.

Reverse-engineering agent for IDA Pro and Binary Ninja that autonomously orchestrates navigation, decompilation, annotation, and parallel sub-agent analysis tasks for binary researchers.

Mobile-friendly, web-based coding agent supporting multiple concurrent sessions and models, designed to autonomously complete programming tasks with multimodal input for developers.

Runs coding tasks locally as single-runtime command graphs, emphasizing measured token savings and higher success rates by understanding repositories before modifying code for developers.

Open-source AI code reviewer for software teams, generating inline comments backed by code evidence to flag bugs, risks, and style issues directly in pull requests.

Terminal-based AI pair-programming agent distributed as one fast-starting static binary, supporting multiple models and local retrieval with sandboxed execution for safe coding.

Provides an open-source operating layer that turns a single CLI into coordinated multi-agent teams, helping businesses automate engineering, operations and collaborative workflows with autonomous AI coworkers.

A locally run personal AI agent that researches and organizes files, writes and reuses its own tools, repairs itself, schedules tasks, and supports voice interaction for individual productivity.