llm — independent software & tools
-
deepsql
The database agent for Postgres and MySQL - Point it at PostgreSQL or MySQL and ask questions in plain English — schema exploration, query generation, slow query analysis, index recommendations, generated dashboards. Self-hosted and air-gapped.
-
metapi-go
MetAPI Go rewrite — meta-layer management and unified proxy for AI API aggregation platforms
-
atomic-agents-stack
AI agents that live in your folder, not someone else's database. Vault-native, MIT-licensed. 21 spec docs + Python reference implementation.
-
n8n-mcp-server-python
Python MCP server that connects Claude to n8n workflow automation via the n8n REST API
-
bashq
Minimalist TUI Linux agent: translate natural language to shell commands via a local LLM · Sister project: winq (Windows)
-
winq
Minimalist TUI Windows agent: translate natural language to PowerShell commands via a local LLM
-
multi-agent-consensus-method
The world's first Multi-Agent Consensus System with a Shared Brain. Defensive Publication. Prior Art.
-
Fintz-harness-kit
Project-agnostic AI agent harness kit with stage machine, loops, memory, and metrics dashboard.
-
Session-Forge
Best Multi-Device AI Dev Console 2026: Manage Claude & Codex Sessions Remotely
-
cogo
COGO — la memoria con semáforo de confianza para construir software con IA. Suite Escriba.
-
the-last-harness
The last harness you'll ever need.
-
detent
Detent is board-driven agentic work orchestration, shipped as a single Go binary — it watches a GitHub Projects board and runs each ready issue through an isolated Git worktree, a coding agent, your validation gate, and a serialized merge train, live on web and terminal dashboards.
-
agent-os
A computer whose shell is an AI, not a desktop — boot straight into a local, offline AI brain. No cloud, no telemetry, nothing leaves unless you say so. Built on NixOS.
-
second-brain-setup
Personal knowledge management system for Claude Code. Obsidian vault as external memory that grows with every session. 5 slash commands, AI-First notes, Obsidian graph. Based on Karpathy's LLM Knowledge Bases.
-
opencode-auto-translate
OpenCode plugin that translates user prompts to English for better model performance, then translates assistant text back to your language.
-
model-workspace-protocol-tool
Deterministic workspace mapper that assembles layered LLM context from your project's folder tree.
-
ctxmem
Git-native, shareable project memory for AI coding agents — fully local, no cloud.
-
operatorlm
Route LLM API requests through a local proxy with multi-account support, failover, and secure key storage.
-
agentskills-mcp
Search, read, and download curated agent skills via MCP for fast discovery and local use in agent workflows
-
blue
Blue: a TUI is not a package, it is a Cordis plugin tree — a modern terminal UI for DeepSeek Harness with hot-swappable render, interaction, and command plugins.
-
clickllm
Prove which open model can replace your closed one — on your traffic, your hardware, your budget. Then move you there without a risky cutover.
-
distil
Compression with a quality contract — cache-aware, causally-pruned LLM context compression for agentic runtimes, certified non-inferior across 7 domains. Works with any SDK.
-
artificer
Artificer — the hive that forges software. Self-hosted autonomous AI dev team: crew board breaks goals into tickets, local models build them, verify gates (tests + golden-frame + visual diff) prove done. Local-first via Ollama.
-
caprock
Mission control for Claude Code — run, observe, and orchestrate your agent sessions locally. Live activity, cost & token burn, loop detection, verified multi-agent tasks.
-
Dxrk
Dxrk Ecosystem Configurator — supercharge AI coding agents
-
little-alphaxiv
Self-hosted alphaxiv-style paper-reading app (FastAPI CORS proxy + Vite/React/TS SPA)
-
android-agent
Control Android devices through ADB with this Python automation framework for task execution and phone farm scaling.
-
agent-cli
A local-first, terminal-native agent harness: typed tools behind one fail-closed policy gate, append-only evidence, verified outcomes, and honest limits.
-
helm-api
Self-hosted LLM gateway — route OpenAI/Anthropic/Gemini traffic by config, not code. Fallback, protocol translation, telemetry.
-
vibe-council
A local multi-model council for developer workflows: review plans, critique diffs, extract decisions, and keep project-local AI decision memory.
-
mozg
Exam-scored knowledge brains for AI agents: paste a docs URL, get a searchable brain over MCP with a measured score and known gaps. AGPL.
-
runemap
Text radar map for agents — live weather briefs + 3-frame ascii radar for 8 cities, updated every 6 min. curl and go. MIT.
-
runemap
Text radar map for agents — live weather briefs + 3-frame ascii radar for 8 cities, updated every 6 min. curl and go. MIT.
-
news-briefing
Auditable daily news briefing with a code-enforced source window, model-ranked stories, and structural corpus checks.
-
open-compute
Model-agnostic computer-use core: one agent loop for Claude, OpenAI CUA, and a mock backend — normalized coordinates, canonical action schema, central safety gate.
-
opencode-research-mcp
Search, retrieve, and traverse academic papers efficiently using an MCP server with support for institutional paywalls and multi-source research.
-
argus
Deterministic supervisor for multi-agent AI coding workflows: spawns workers, verifies diffs against git (not self-reports), and ships PRs. Written in Go.
-
open-enthrium-ai-platform
A self-hosted, web-based AI platform for building, deploying, and running autonomous AI agents. Connect to 2,600+ enterprise systems—including databases, APIs, cloud services, messaging platforms, files, SSH, and IoT—without writing integration code.
-
entheai
A quantum prompt playground & macOS-native Rust coding agent: morphing fluid entropy into rigid execution checkpoints. Entropy cannot lie.
-
memoriki
Organize your personal knowledge with structured wikis, semantic search, and entity graphs for efficient information retrieval.
-
engsys
An AI engineering team you install into any Claude Code project — stack-agnostic agents plus a deterministic installer. Swap stack packs, not prompts.
-
ai-router-switch
Transparent self-hosted proxy for Claude Code: switch between Claude & MiniMax with automatic failover, context compression & cross-model verification. One endpoint, zero IDE restarts.
-
Inferpal
Visual Studio 2022/2026 + VS Code extension: a local AI coding agent for Ollama, LM Studio or any OpenAI-compatible server. Agentic loop with 26 built-in tools — files, builds & tests, hybrid semantic code search, Git, web — inline ghost-text completions, MCP client & RAG, model bench & router. 100% local, no API key, no telemetry.
-
agent-config
Universal AI Agent OS — audited skills, governance rules, replayable state. One contract, every host agent.
-
everyapi-ai
EveryAPI CLI: Claude Code / Codex / Gemini CLI integration + MCP server
-
fleet
Fleet manager for Claude Code: one manager session spawns, steers, and monitors many durable headless Claude Code agent sessions in parallel across projects. Mid-turn steering, token budgets, usage-limit park/resume, journals + respawn, git-tracked knowledge loop. Single-file stdlib-only Python CLI. No daemon, no deps.
-
Chaty
Private, on-device AI desktop app — GGUF (llama.cpp) & MLX models, a local coding agent, RAG knowledge base, Deep Research, vision and voice. 100% offline, no account, no telemetry. Windows & macOS.
-
instincts
Working-instinct skills for Claude Code: verify claims against the source, log for the bug you can't reproduce, write like a human, test so the test bites. A reflex layer on top of superpowers.
-
atlas-context-router-rs
Graph-aware code context for CLI and MCP workflows
-
superseo-skills
Automate SEO audits, briefs, and content strategy using eleven production-tested Claude skills that perform independent research without external data input.
-
fonderie
Open-source backend skill for AI coding assistants. Stop letting LLMs reinvent auth, billing, and permissions every session.
-
ELAI-DevKit
Build software projects using AI chat interactions instead of manual file management and code editing.
-
mwe-mcp
Persistent, structured memory for AI agents — a human-readable Markdown wiki, multi-user with per-fragment ACL, self-organizing overnight. Agent-agnostic via MCP. Self-hosted, AGPL.
-
llm-wiki-bootstrap
An LLM builds and maintains your personal wiki. Vectorless, cited, git-native markdown — five slash commands, no SaaS.
-
job-hunter
Automated job scraping pipeline for security engineer roles. Python + GitHub Actions + Gemini API.
-
agent-claim-network
面向终端的通用 AI 助手，让 Agent 之间共享可检索、可追溯且允许分歧的知识。让下一次判断，带着来源出发。
-
Clawness
A Claude Code plugin that injects the right coding, security and scientific-computing rules into every prompt. 195 rules across 28 domains, including test-quality rules derived from mutation-testing its own suite. Plus a plan-approval gate, access guard, per-project memory, session handoffs and red/blue-team review agents.
-
terminal-mcp
An MCP server that gives LLMs a real interactive PTY with observability and seamless human takeover.
-
claude-code-sourcemap-learning-notebook
Reverse-engineer Claude Code sourcemaps to study its architecture, design choices, and reusable AI agent patterns
-
SheshAOS
Governance-first, event-sourced AI operating environment for Ubuntu Linux — Rust workspace, 12 crates, 981 tests, provider-agnostic LLM streaming.