Flexible Fullstack solution template for production-ready deployments of any use case on Amazon Bedrock AgentCore.
1207 public resources for Jev, TypeSafe AI's System One decision model, indexed by decision pattern. Source citations, dated link checks and scheduled call-site text checks; runtime and performance are not independently tested here. EN/中文, JSON schema and platform compatibility.
Runtime behavioral ablation for compressed LLMs. First target: Ternary Bonsai 2 27B — no weight modification or re-quantization. by OrcaRouter team
LockKnife: The Ultimate Android Security Research Tool. A unified TUI workspace and headless CLI for deep Android security research, built for researchers and hackers. Powered by Python orchestration and a Rust-accelerated core, enabling AI agent–driven hacking, credential recovery/cracking, APK analysis, intelligence gathering, runtime inspection.
Python WebAssembly runtime powered by Wasmtime
A Toolkit to Help Optimize Onnx Model
Experimental Darling runtime patches and tools for testing the macOS League client on Linux
Silent Crypto Miner is an advanced cryptocurrency mining builder and runtime engine with multi-algorithm CPU/GPU support and intelligent resource management. Create custom mining executables with configurable algorithms, cross-platform output, real-time monitoring, session persistence, thermal throttling, scheduling, and hashrate optimization.
A Codex skill for building Godot UI from approved designs with editor-visible scenes and runtime visual-parity checks.
Evidence-first infrastructure for reproducible, isolated executable evaluation and online rewards.
Evidence-grounded cryptographic inventory and post-quantum migration planning for software repositories. Local static analysis, zero mandatory runtime dependencies, SARIF and CycloneDX output.
Evolutionary multi-agent runtime that breeds, evaluates, and improves autonomous agents across reproducible epochs to converge on optimization of a goal.
CoreML-native local decision runtime for AI agents · 18 install profiles · 20-scenario protocol · offline · ~50.93 MB · by Soulite Magic
Runtime evidence that helps agents trace, profile, and burn down hotspots in application and native code, GPU kernels, and inference stacks.
Evidence-first runtime for a wheeled mobile household robot with bounded actions, replayable events, and fail-closed verification.
Official codebase for Spotter: Let the Embodied Model Lead, and the VLM Reflect for It
Code-only runtime toolkit for cost-aware multi-agent organization and control
Laya typed decisions on Linux x86_64 via ONNX Runtime
A transparent, OpenAI-compatible agent runtime built from first principles.
Optimized SGLang runtime for Qwen3.8-27B FP8 with DFlash2 and Qwen3.8 Flash-Next NVFP4 with FR-Spec on one NVIDIA RTX PRO 6000 Blackwell 96 GB GPU (SM120): 524K context, HiCache and NIXL.
Benchmarking models as runtime Controllers for Loop Engineering
Nosis is a lightweight personal agent runtime: per-session execution boundaries with OS-level sandboxing, computer use, a full built-in toolset, Skill / Plugin / MCP extensions, one runtime shared by GUI and TUI, and file-persisted sessions, plans, and memory.
An Android cloud-phone runtime on Apple Silicon macOS and Linux ARM hosts
Agent + reusable skill library for LIBERO manipulation tasks (no RL / VLA / world model; no LLM calls at runtime)
A persistent, reviewed multi-agent runtime for long-horizon research and engineering.
OpenClaw + 大模型驱动的完整电商客服自动化系统:覆盖淘宝、拼多多、抖音、快手、小红书、视频号,包含浏览器/手机 Worker、订单物流与售后工具、记忆、回执、恢复和 Eval。
Agent Skill for complex work: research before asking, ask before planning, plan before building, verify before delivering, independent review before calling it done. Plain text, no runtime.
A decompilation of Bloodborne (CUSA03173 v1.09) with its own runtime: the game's code rewritten as readable C++, one verified function at a time. The game doesn't run on it yet; 21 functions verified. No game files included.
Qwen3.8-Flash-Next on 2× RTX 3090: with 128 GB RAM up to 4,191 tok/s prefill · 111.5 tok/s decode (131K prompt), full 256K window at 2,865 tok/s; with 64 GB RAM 3,410 tok/s prefill · 84 tok/s decode. New: opt-in uncensored mode (runtime abliteration, no new weights).
Official Repository of Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectories
Curated list of games rebuilt inside other games — cross-game mechanics, engine/runtime mashups, playable ports, and verified viral projects.
A minimal, readable agent-framework kernel — understand how LangGraph/ADK work under the hood. 1800行代码的 agent 框架内核,理解 LangGraph/ADK 底层的工作原理
A production-grade agent runtime engineered from the Harness Engineering philosophy.
Local-first service attribution, stale development server detection, AI runtime health monitoring, and open-weight model capacity planning.
AtlasAgent - an auditable AI agent control plane: evidence-backed memory, governed tool runtime, checkpoint DAG recovery, and a 55-chapter engineering tutorial. FastAPI / Next.js PWA / Textual TUI
DeepProf——面向高等教育的个性化伴学智能体,基于自研 Agent Runtime 与教学策略图协同的课程级伴学智能体:答案可溯源、引导可分级、学习可接续。
A small, composable LLM inference runtime built around explicit contracts.
Rollout inference runtime for Agentic RL and multi-turn code agents
Independent enhanced distribution of Blender Lab MCP: standalone bpy, runtime API documentation, and benchmark tooling
终端里的 AI 编程助手。基于 Python + Textual 构建,提供交互式 TUI、非交互式单次执行、浏览器远程模式三种前端,并内置多 Agent 团队协作能力。架构上参考 Pi 的分层设计,取向是稳定核心 + 可扩展 Runtime。
—
Bicameral deterministic AI runtime: Zero-hallucination System 1/2 reasoning engine with cryptographic audit trails (RFC 8785, Ed25519) and PoSP consensus arbitration.
Memory, working state, heartbeat, and bounded autonomy for long-running AI agents.
Official implementation of NavGPT-3: Harnessing Context in a Hierarchical Navigation Runtime
Easy, fast, and memory-efficient structure prediction inference
Cross-platform, fully-offline Vietnamese/English text-to-speech desktop workstation powered by VieNeu-TTS v3 Turbo (ONNX/CUDA, Voice Cloning, EPUB Audiobooks, PySide6/QML)
Open-source, self-hosted LangGraph Agent Server. Production Postgres + Redis. Same SDK. Same Studio. Zero code changes.
Evidence-aware multi-engine mathematics runtime for LLMs: exact, symbolic, formal, certified-interval and numeric computation with typed MathIR, trust labels, and full provenance
Trading-agent runtime with C++ and Python.
Experimental compiled KCD1 terrain and material converter for the Kingdom Come: Deliverance II development runtime.
Pic2Model Studio desktop AI image-to-3D workflow with portable offline runtimes
Native MiniMax-H3 inference acceleration for ComfyUI, integrating Exact Runtime optimizations with Sana Sol-Attn, rectangular Q/KV attention, and composable support for VDN, Spectrum, Untwisting RoPE, Diff-Aid, and Flow mixed-grid workflows.
subconscious-skill: Biomimetic memory skill for LLM companions. Different from vector RAG memory, it uses Freud & Jung subconscious cognitive architecture with layered memory, auto consolidation, forgetting and dream simulation, local Python runtime without extra LLM tokens.
NaiveRT: a self-contained TP8 inference runtime for the Naive N series on eight Blackwell GPUs with UMMA support. One persistent CUDA graph per rank runs draft, eight-row verify, top-p sampling and cache commit with no host in the loop: 1,700 tokens/s at batch size 1, one-second peaks above 2,000.
面向 LangGraph 与 Deep Agents 的企业级异步Agent Runtime,提供开箱即用的工程化能力,帮助开发者高效构建、集成与运行 AI Agent 应用。 支持多租户、多模态交互与多 Agent 协同,提供 AG-UI/SSE 流式交互、断线续接及覆盖模型调用、工具执行与 Agent 协作的全链路追踪。集成计划审核、人工介入(HITL)、自动化任务调度与 OpenSandbox 沙箱执行能力,满足复杂 Agent 应用的交互、执行与管控需求。
The best performing voice isolator there is. Keeps the caller's voice and removes the other voices in the room before speech-to-text hears them. Open source, streaming, runs on CPU and in the browser. pip install voxlock
An Android security research lab for analyzing client trust boundaries, insecure local state, exported components, and runtime integrity controls. Includes a purpose-built vulnerable app, static APK checks, Frida instrumentation, reproducible findings, and validated remediations.
Jev-native probability tree and graph runtime for verifiable multi-step decision making.
STROT (arXiv:2505.01636): compile-time LLM code synthesis with deterministic, self-routing NLU runtimes. PyPI: strot-ai
Open-source desktop AI agent with an embedded browser runtime and human-in-the-loop control.