Tag
The post argues that Anthropic's repeated restrictions on reverse engineering tasks in Claude Code (and Opus 5.5 blocking game-cloning work) is yet another reason to self-host local LLMs, warning that blanket frontier AI policies ensnare otherwise legitimate peripheral behavior.
rauchg argues for owning your AI stack by running local models and embracing open source, urging users to control their harness, data, connections, runtime, and agents rather than letting others resell their intelligence.
A user shares a detailed setup guide for running DeepSeek-V4.1-Flash-1017 (~35 tok/s) and Qwen3.8-Flash-Next (~130-140 tok/s) on older consumer hardware (RTX 3090 Ti, 128GB RAM, Threadripper 3990X), including an NVIDIA firmware-throttle unlock via an open-source kernel module and an Xolotl inference engine build.
braindead-local-intelligence is a simple podman-based project that sets up local AI on a Linux machine: an OpenAI-compatible LLM server, an image generation server, SearXNG web search, and Open WebUI behind a single address, loading only one model in VRAM at a time and unloading it when idle.
This tutorial explains how to use Cloudflare Tunnel (cloudflared) to turn an old computer at home into a publicly accessible server without a VPS, public IP address, or port forwarding. It covers getting started with Quick Tunnel, long-term deployment with a named Tunnel, and common pitfalls.
NetBird is an open-source platform that combines WireGuard VPN, peer-to-peer connections, and centralized access management to simplify creating secure private networks, now also introducing a beta Agent Network for identity-aware access control for AI agents.
I've open-sourced a self-hostable, end-to-end encrypted relay server solution (relay-kit) that can be deployed with one click using just a Cloudflare account. It supports tunneling any HTTP/WS service and is built around a zero-dependency encrypted relay that blindly forwards traffic.
Tangled announces itself as the first truly federated code forge built on AT Protocol, allowing self-hosting of Git, CI, and UI components while collaborating across instances, alongside a major UI redesign and a rearchitected backend with a Rust-based in-memory indexer (bobbin).
sitesolide 是一个免费开源（MIT）的部署工具，可让你用 Claude Code 或 Cursor 等工具生成的应用一键部署到自己的服务器上，自动配置 HTTPS 和独立域名，并支持应用间隔离与访问控制。
A Lobsters user asks for recommendations on a shared password manager suitable for non-technical family members across long distances, noting concerns about 1Password and alternatives like KeePassXC and Syncthing.
OneDeploy 是一款开源应用一键部署服务，用户可从 105 款应用中选择并在几分钟内于专属服务器上运行，按小时计费并封顶月费，无按人头收费，自带 TLS、域名绑定和每夜加密异地备份。
A practical guide to running a lightweight web server on NetBSD 11 using the stock bozohttpd daemon, with HTTPS set up via acme-client and Let's Encrypt, aimed at small-scale or homebrew projects.
作者分享如何用一台12年前的办公PC运行本地脚本（IMAP邮件轮询、Tesseract OCR、Piper TTS），把Telegram机器人例行任务从昂贵的LLM API调用转为零token的本地处理，显著降低成本。
博主分享如何利用 Caddy 的 forward_auth 指令，仅向公网暴露私有自托管服务（如 Immich、Nextcloud）的分享链接，其余请求一律返回 404，从而在无需 VPN 或账号的情况下提升安全性。
A user describes how their AI agent autonomously handled a PHP vulnerability across their self-hosted home network — patching updated services, stopping unpatched containers, and setting up a watchdog to apply fixes as they became available.
开发者发布免费工具 ModelAnalysis.si，可跨 450+ 模型计算 API 调用与自托管（RunPod 租用 GPU）之间的成本盈亏平衡点，支持输入/输出 token 量与缓存命中率调节、VRAM 量化规格估算，以及基于实时定价和 ELO 排名的帕累托前沿分析。
作者分享了使用 NixOS 部署 FreshRSS 等自托管服务的体验，说明声明式配置如何大幅降低运维和重建服务器的心理负担与工作量。
作者放弃了 Meta 的 Muse，转而使用开源、可自托管的个人代理 Comma（AGPL-3.0），因为其代码可审计、数据可本地运行，无需信任 Meta 的隐私承诺。
A free browser-based LLM cost calculator lets you compare local GPU inference (with RTX 4070 Ti Super / 5090 presets) against cloud API pricing, highlighting that cached-input token rates — often guessed at by other calculators — most determine where the local-vs-cloud crossover sits.
OneDeploy 是一个 Show HN 发布的托管开源应用平台，支持 88 种应用，提供按小时计费（月度封顶）、每日加密备份、自动 TLS 证书和区域选择，让用户几分钟内在自己的服务器上部署团队工具。