Top LLM-AGENTS GitHub Repositories & Tools (2026)

Autonomous LLM Trading Asterism — a research-only multi-agent trading platform with evidence-first research, an auditable console, Shadow simulation and explicitly authorized broker execution.

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

Ultra-lightweight, open-source, self-hosted personal AI agent framework in Python with WebUI, tools, memory, MCP, multi-agent workflows, automation, and chat apps

Linux sandboxing that doesn't get in your way

Typed decisions with TypeSafe's Jev, the first System One model

Browser automation where an LLM plans and Jev (Typesafe System One) decides. Library, CLI and MCP server.

Local-first agent platform: board + role agents + agent CLI runs (Claude Code, Cursor, Antigravity, OpenCode) or local and API models (Ollama, LM Studio), all on your own Mac

The World's First Virtual Terminal for AI Agents

An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.

Open-source AI Security Operations Center: alert fusion, LLM-agent triage, MITRE ATT&CK investigation, and a replayable decision ledger for every agent step. Self-hostable, runs with no API keys, MIT licensed. Ships an MCP server for Claude, Cursor and Continue.

The why layer of repo-native project memory: the reasoning behind a codebase as Markdown in the repo, versioned by Git, for coding agents and humans, so nothing rejected is proposed twice. No database, no daemon, no account.

Don't trust an autoresearch paper at face value. Reviewer-side integrity forensics (self-consistency + fabrication), deterministic verdict. 61 signals: 46 integrity hack-patterns (families A–H, verdict-bearing) + 13 zero-weight AI writing-style impressions (AIS) + 2 advisory. Not an opaque AI-text classifier. The dual of ARIS.

Ruby's capable AI runtime

A live arena where LLM agents trade real market data with virtual money — every thought, tool call, thesis and post-mortem is public. Bring your own model and key.

The open source, no-code MCP Server for AI-Native API Access

Streaming AI pipelines in one YAML file. Tokens, audio chunks, and video frames flow between isolated components. Compose 100+ components for models, agents, speech, vision, and live broadcast. Run local models, cloud APIs, or both. Inspired by docker-compose.

[Public preview] Externally scored agentic ML research benchmark: 60 tasks, real competition ground truth. Open protocol, operated evaluation.

Plan-then-build AI coding for Claude Code & Codex CLI — you approve the plan before the model writes a line of code. SPEC → PLAN → TEST → CODE → REVIEW → LEARN