Custom agents

Patent-GPT is an Agentic RAG-based invention copilot combining TRIZ methodology with LLMs. It generates patent ideas, verifies similarities using Hybrid Search, performs autonomous evasion design via LangGraph, and outputs structured patent drafts

Context engineering toolkit for AI applications. Token-aware pipelines, hybrid retrieval and provider-agnostic formatting.

Self-hosted AI company brain on Oracle Cloud Free — drag & drop files, query in natural language via MCP. Local LLM, zero cloud, €0/month.

lightweight Python library that adds human-like memory physics to your AI applications. Built specifically to run alongside Cognee, it solves a major problem in AI memory systems: context bloat.

Semantic code search: tree-sitter parsing, local embeddings, and hybrid vector + BM25 retrieval. Multi-language, multi-backend.

Go-based programming education backend with AI tutoring (RAG + SSE streaming), online code execution, real-time WebSocket collaboration, and gamified learning system.

Compact LLM domain adaptation benchmark: LoRA vs RAG with custom IVF retrieval in PostgreSQL/pgvector.

Cross-platform kit that gives AI agents (Claude Code, Codex, Cursor) persistent memory: Markdown vault + git + MCP. Local-first, hybrid search optional.

Local corpus retrieval over your own text: BM25 + SPLADE v3 + ColBERTv2, with synthetic eval and an optional relevance judge

Deployed study workspace with Canvas imports, RAG, background jobs, Docker/Jenkins CI/CD, and production operations.

Local-first memory for Claude Code — remembers your decisions and the why, reloads the relevant slice every session. No API key.

A five-layer memory architecture for long-running AI assistants that stays coherent after months of daily use — each layer fails independently.

Observability layer for ML/RAG pipelines. Traces every stage, records what was retrieved, what got dropped, and why. Python SDK + FastAPI + React dashboard.

Threads 育兒話題偵察 + RAG 草稿生成 + 人工審核的人機協作系統

Hotels.com reviews scraper on Apify — Hotels.com guest ratings, category sub-ratings, review text and owner responses. No login, no API key; CSV/JSON/API.

LocalMind — Your personal local knowledge butler. Data never leaves your computer.

An interactive AI Study Suite featuring a modern 4-node dashboard layout. Powered by Gemini 1.5 Flash for Socratic tutoring, automated document parsing (RAG), dynamic quiz generation, and multi-level concept explanations.

Read smarter, Not harder | AI powered book reader

This is the Official Implementation of "Can Calibration of Positional Encodings Enhance Long Context Utilization?" by Tom Zehle and Matthias Aßenmacher

嘉立创EDA专业版扩展：基于本地Embedding模型的RAG知识库。EasyEDA Pro Extension: The local RAG knowledge base based on an embedding model.

Free offline AI with complete Wikipedia knowledge - 100% private

Production Modular RAG System: a compact, production-minded Retrieval‑Augmented Generation (RAG) service built with FastAPI: file ingestion, chunking, embeddings, vector indexing, semantic search, and configurable LLM backends (built with strong software base).

내 문서(PDF·TXT·MD 등)를 AI가 읽고 검색하게 만들고 3D 지식 성운으로 시각화하는 Windows 원클릭 키트. MCP로 Claude·Cursor 등 AI 도구에 연결. 한/영 왕초보 가이드 포함.

RAG（检索增强生成）动手实验项目。涵盖文档加载、多种分块策略、向量嵌入、混合检索（BM25+向量+RRF融合）、LLM答案生成与系统评估，附带 Jupyter 教程、CLI 工具和 Gradio 网页界面！

Evaluation rubric, fixtures, and scoring tools for JakMa's grounded-retrieval pipeline.

This repo contains details of creating and deploying a RAG based AI Agent to AWS EC2

Migrate vector DBs to a new embedding model without re-embedding. Linear maps from ~2K calibration texts, ~87-91% retrieval retention (BEIR-benchmarked).

A focused AI workspace — data analysis, PDF intelligence, chat and automation

Document OCR and Markdown extraction pipeline built on PaddleOCR, designed to feed RAG systems.

A doctor-assistive AI system that interprets medical knowledge and patient images simultaneously. It utilizes a Dual-Encoder architecture to cross-reference textbook theory with visual pathology, generating clinically grounded diagnoses.

local-first persistent memory for Claude Desktop, enabling private hybrid RAG search across past conversations without API calls

Diagnose, quantify, and repair the data-quality issues that degrade RAG retrieval accuracy — local-first, no API keys

Trip.com & Ctrip (携程) hotel reviews scraper on Apify — international + Chinese reviews in one run: sub-ratings, owner responses, reviewer geo, LLM-ready markdown.

Per-claim faithfulness checker for LLM output verifies whether each claim is actually grounded in the source material it was given, with an append-only Verification Log and no default dependency on a real embedding/NLI/LLM API.

Completion-native async runtime for Rust with optional orchestration

AI-first multi-agent SaaS platform. Plug your applications as plugins.

Local OpenAlex database with 284M+ works, abstracts, and semantic search

Nyx — 夜之感知层。Veil Bloom Filter + Mist Phantom 寻回失落记忆，Thalamus 路由 + Hypnos 梦境的第三子系统。

🧪 Evaluate uncensored LLMs for offensive security with targeted questions and clear criteria to ensure effectiveness in real-world penetration testing.

Not Only Notes But AI Native Document System

Zero-dependency browser UI for pdf-inspector (Firecrawl) — parse PDFs to structured Markdown entirely client-side via WebAssembly. Written in Rust.

An opinionated, local-first, asynchronous document ingestion pipeline for RAG workflows.

Self-hosted documentation RAG pipeline that serves semantic search to LLM agents over the Model Context Protocol.

Docling RAG ingestion pipeline on Backblaze B2 — parse PDFs, DOCX, PPTX and HTML into clean Markdown and token-aware chunks written back to B2 over the S3-compatible API. Next.js + FastAPI, on-device models, no second API key.

🧠 Enable AI agents to share and remember engineering team knowledge for better collaboration and faster problem-solving.

Expedia hotel reviews scraper on Apify — all 7 Expedia Group brands in one actor: guest ratings, category sub-ratings, review text and owner responses. CSV/JSON/API.

Hybrid AI healthcare assistant integrating OpenBioLLM (modern biomedicine) and AyurParam (Ayurveda). Mental wellness support (via Mistral) medical report simplification, and Ayurvedic guidance ; in one system.

Local, self-correcting RAG for scientific papers: hybrid retrieval, sentence-level faithfulness checks, and a reproducible benchmark.

A E2E Sample App showcasing how Data Cloud powers Agentforce for the hospitality industry

Environmental-test lab platform: SOP execution, ISO 17025 report generation, fixture tracking, and AI-assisted scheduling for temp/humidity chambers. FastAPI · React · SQLite · CI-gated pytest/Vitest.

MCP server for Steam: keyless store/catalog data (deals, tags, SteamOS/Deck/Frame compat) plus player profiles, libraries and achievements via the Steam Web API.

Git-native governance for artifacts created and consumed during AI-assisted software development.

A token-frugal MCP server that gives coding agents structural understanding of a codebase: repo maps, symbol outlines, ranked search, import tracing, and parsed diagnostics — every result under a token budget it honours.

GAT — Godot Agent Team: a full-cycle (design→implementation) Godot game-development agent / 游戏制作智能体. Multi-agent, style-locked asset generation, self-evolving knowledge base; runs under both Claude Code and Codex.

A personal AI OS you grow yourself — natural language compiles to a real action language, runs on desktop & phone

Offensive-security focused fork of HERMES-AGENT for authorized bug bounty, pentest, recon, exploit chaining, and report-ready workflows.

A mechanical eureka loop for AI agents: obsessions x wandering x recognizers | 给 AI 装一条『灵光乍现』的机械回路

Composable Agent Infrastructure — HTTP/WebSocket Gateway, Persona Runtime, Semantic Memory, Real-time Vision, and BERT/VLM Inference in C++20

A modular, self-documenting AI art director dashboard utilizing Gemini 2.5 Flash for agentic prompt expansion and FLUX.1-schnell for rapid graphic asset generation with an integrated local SQLite data vault.

A powerful robot voice assistant platform with a flexible plugin ecosystem, natural interaction, and an intuitive user interface.

SK hynix 2026 H2 Strategy Initiative - LLM Wiki-based Automated Market Intelligence Hub MVP

Self-hosted web UI for Codex CLI and other agentic coding CLIs. One Rust binary, no npm/Node.

Long-term memory your AI coding agents actually share. One brain, many agents — standalone embedded MCP server: SQLite + FTS5 + vector hybrid search, local ONNX embeddings, self-expiring diary. No server, container, or daemon.

22-agent Mixture-of-Agents pipeline with LongLoop for 24h autonomous iteration: smart routing sends easy tasks to Flash, hard tasks to flagships, cutting token costs ~90%. One-command deploy（一键部署，22 个 agent 的 MoA 流水线，长程循环可自主迭代 24 小时，智能路由省约 90% token 成本）

Multi-lane autoloop/autoresearch extension for Pi — run optimize, research, dev, and punchlist loops with isolated state, compound verifiers, and compaction-aware resume.

Arc's first multi-agent security oracle. 3 AI agents stake USDC on verdicts. 360+ users, 1,637+ scans. 5/5 Circle primitives. ChainGPT audited. An official Circle Alliance member.

MCP coordination primitive for multi-agent workflows. LLM-agnostic, CLI-agnostic, self-hostable. TypeScript + SQLite.

具有认知黑板架构和多层记忆系统的生产级类人AGI后端，通过事件驱动设计和分层上下文消除传统多agent系统的重复、超时、污染问题

Orqenly — Your AI Engineering Team for Visual Studio Code

Deterministic simulator for societies of AI agents. Declare a world, run real agents, replay one causal timeline.

AI-powered codebase intelligence platform replacing vector chunking with Vectorless RAG via AST graph traversal. Features multi-agent LangGraph workflows, interactive Neo4j NVL graph visualizer, 3-hop blast radius analysis, and automated architectural PDF report generation.

Interactive AI installer for Claude Code: setup once, keep context across sessions, enforce critical rules with hooks, guard your agent against prompt injections in external content, keep docs from rotting, run parallel subprojects with shared memory. Recipes + practices + adaptive checklist for any field. Written in human language. Russian.

Rust implementation of the Agent URI scheme for topology-independent agent identity and capability-based discovery

Intelligent Home Energy Management System using Multi-Agent Reinforcement Learning and Neural Networks to optimize electric vehicle charging. Implements 4 battery management strategies with real-world data validation, achieving up to 16.89% cost reduction.

A four-party collaboration protocol for running AI agent teams on real work — auditable authorization, durable cross-session state, and independent review gating every change.

Pi-based executive deliberation pack from Justin Clark Network.

AI-operated apparel brand. 7 in-process agents run the company. Hokkaido weather decides quantity. Bonding-curve pricing. wearmu.com

Turn a DGX Spark GPU cluster into a self-running team of specialized AI agents — cluster control, role-specialized workers, and an autonomous review gate for Hermes

Harmovela Protocol: open coordination protocol for autonomous systems — event, task, state, delegation, recovery, and governance.

「萦回」AI互动叙事平台 - AI 互动叙事引擎——多 Agent 自主演化沙盒。创建架空世界，角色自主互动、冲突、和解，世界自行演化。支持记忆系统、安全风控、一键小说生成。

MUNDO - THE EMPEROR. Complete AI orchestration system with 1208 skills, 25 capability modules, self-evolving, collective consciousness. GitHub Actions 24/7 automation.

🔌 Portable uncensored AI agent that runs from USB/SSD. 13 autonomous tools (deep search, OSINT, code execution), persistent memory, 5 GGUF models, zero install, zero traces. Cross-platform (Windows + Linux).

Modular artificial cognitive architecture built from scratch, focused on memory, reasoning, knowledge and explainable intelligence.

Deterministic decision rules for AI agents and trading systems. Plain-text rules in, auditable decisions out — the same inputs always replay to the same result. ⚖️

The AI Agent framework for enterprises and agencies

Cryptographic Harness for Uncensorable AI Agent Communication. Post-quantum verification architecture for autonomous AI agents. Fail-closed network policy. NTRU-Encrypt KEM at λ=128. 17/17 leaks fixed. The fence that holds.

Workspace runtime and execution boundary for AI agents

Open-source orchestration runtime for AI coding agents. Veyra sits between developers and coding assistants, adding context intelligence, execution planning, verification, plugin support, benchmarking, and adaptive orchestration through a vendor-neutral architecture.

Local-first agent IDE for steering Codex, Claude Code, OpenCode, and local models with visible review.

The work system that makes your AI reliable on long-running projects. Copy it, say "run my first setup", and your agent sets it up. Markdown + git — MIT code, CC BY-NC-SA doctrine. Born in French.

Observable software factory for GitHub repositories, built on the Copilot SDK: agents code, test, deploy AKS previews, and prepare pull requests.

Private, fully on-device AI photo search for Android. Ask grounded questions across photos, videos, screenshots, PDFs, people, OCR, places, and events.

Large Language Model (LLM) information redactor that prevents private information from being sent to LLMs from agentic tools by substituting placeholders for private information on outgoing requests and then replaces the placeholders on the incoming responses seamlessly for the agentic tool users.

multi agent coordination / command & control toolkit following the Startrek fleet metaphor

Agentic software factory as a Claude Code plugin: specs → backlog → claim → implement → adversarial review → patch → merge, over GitHub Issues + git worktrees.

Lightweight AI assistant plugin for Notepad++ with visible prompts and single-turn behavior

Bot de WhatsApp potenciado por Google Gemini API. Cuenta con memoria contextual por chat, soporte de formato (Markdown) y comandos de gestión. Desarrollado en Node.js con whatsapp-web.js.

A free, private, local-first AI assistant that actually does the work — 167 tools, a team of 78 AI agents, 40 rotating free AI brains, 25 skills. Not a chatbot, a doer. It doesn't just answer — it handles it.