Developer Corner

A distributed SQL database explains its storage, replication, and search machinery through live browser diagrams and a PostgreSQL-compatible interface.

Orblit puts game logic and interface in Flutter, while a C++ entity-component core and Filament handle the engine work underneath.

Bivdi proposes a capability-secure, object-centric operating system in which programs receive only the authority explicitly handed to them.

Engineer and maintainer Xianpeng Shen presents tools for linting, commit standards, CI/CD automation, identity infrastructure, Jira administration, and encrypted Mac backups.

Datasint creates 3D-rendered, labelled training images for rare or temporary road signs and other hard-to-photograph objects.

A command-line tool analyzes coding-agent sessions and repository activity, then places builders on a public ranking.

A continuously updated security-intelligence feed for exploited vulnerabilities, live incidents, ransomware activity, and critical CVEs.

A Unity lighting system for dynamic global illumination across mobile, standalone VR, and PC, with a browser demo.

Ante is an open-source macOS terminal that tracks Claude Code, Codex, OpenCode, Pi, and ordinary shells across working, waiting, and idle states.

A Spanish-language weekly bulletin tracks breaking changes in Amazon’s Selling Partner API models and explains what developers should inspect.

APIBreak compares a vendor's OpenAPI specification with a pinned revision and reports only breaking changes affecting your declared endpoints.

A browser-based robotics simulator that turns a plain-language description into a furnished 3D workspace with real-world dimensions.

An open-source command-line recovery tool that bootstraps fresh Macs and Linux machines or repairs broken coding setups.

A self-hosted platform for turning 360° camera footage into privacy-reviewed maps, walkthroughs, and place pages.

A local Windows tool for chatting with AI models from VS Code while monitoring spend, latency, tokens, and secret redaction.

Bosonuum computes and verifies the physics of superconducting devices before fabrication, then records what passed and what remains unresolved.

An engineer’s workshop documenting four personal automation tools for money, licensure paperwork, gym training, and cooking.

A browser-based Rust analysis tool that reads public repositories and explains runtime behavior, from spawned processes to ignored checks.

A browser-only collection of DevOps converters, calculators, builders, and snippet generators.

An open-source editor for the Aleph One engine that brings Marathon map-making, assets, physics, terminals, compilation, and live playtesting into one workspace.

An open-source AI avatar system that carries one identity, memory, and workspace across a console, Roblox, X, Discord, and a terminal.

Earshot runs synthetic test calls against IVRs and contact-center phone trees, checking prompts, spoken values, delays, and routing behavior.

A volunteer-run Vancouver workshop where members share quadrupeds, robotic arms, teleoperation rigs, edge computers, simulation tools, and a padded test cage.

A Greek software studio building open-source tools for running coding agents, session memory, and children’s learning companions on local hardware.

A developer-facing exercise catalogue with short videos, stills, hand-checked classifications, bilingual metadata, and substitution links.

A macOS application for running multiple independent local PHP servers, each with its own PHP, MariaDB, Apache, database, port, and configuration.

A proxy that reserves each agent's estimated call cost against a cap before sending the request to an AI provider.

A desktop SQL client for PostgreSQL, MySQL, and SQLite, with persistent worksheets, database browsing, search, and readable results.

An independent, source-linked index of 154 Jev applications, browser agents, games, integrations, SDKs, and experiments.

Jev is a hosted AI model and API that turns messy text into typed choices, scores, and yes/no probabilities for software workflows.

A self-hosted gateway centralizes MCP servers, refreshes upstream OAuth tokens, and gives Claude Code, Cursor, Codex, and claude.ai one URL to use.

An open-source Mac workspace that puts Claude Code agents, a browser, inbox, calendar, and indexed knowledge in one window.

A native macOS menu-bar app that tracks Codex and Claude Code quotas, reset times, reading age, and background agent processes.

Independent tutorials explain Jev, a typed AI model that turns unstructured state into parallel probabilistic decisions instead of prose.

An open-source command-line tool records commits, AI sessions, bugs, and one-sentence decisions, then turns the selected material into a shareable project page.

Midas offers operating data, production code, legal archives, and a live agent-training environment drawn from two businesses.

OfficeGPU is building private, local AI infrastructure that routes workloads across the NVIDIA RTX workstations a small office already owns.

A research site about moving probability distributions through the geometry of optimal transport.

A visual editor and PS1-native toolchain for making PlayStation 1 games with Lua scripting, scene tools, asset imports, and hardware-aware build workflows.

A managed venue-intelligence service collects authorized posts, extracts structured records, and attaches citations to every asserted field.

A searchable inventory of VPS, cloud, bare-metal, GPU, and PaaS offers, available through a web query, API, CLI, or MCP.

An open-source workflow engine for AI agents that enforces dependencies, schema-checked outputs, human approvals, and tamper-evident evidence.

A free, open-source local workbench for preparing data, fine-tuning language models, testing adapters, comparing results, and exporting GGUF files.

A personal site collecting open-source machine-learning projects, technical writing, and work on production systems for embeddings, retrieval, and language models.

A local-first macOS client for HTTP and GraphQL work, with Jaeger tracing, response-driven variables, schema-aware editing, and request timelines.

Spender is a free, open-source macOS app that gathers LLM API spending from providers including OpenAI, Anthropic, DeepSeek, xAI, Kimi, Qwen, and OpenRouter.

Vesqo reads billing, product analytics, code, errors, support, and CRM data together to explain what changed, which customers were affected, and what to do next.

A self-hosted debugging tool that follows one customer, order, or invoice record across services and identifies where its data changed.

ado wraps existing coding agents in a workflow for specification, specialist review, budget controls, human approval, and audit records.

A self-contained local aircraft radar that receives, decodes, displays, and remembers nearby ADS-B traffic without a cloud account or subscription.

A specialized auditing engine that tests pricing data for lockstep movement, machine-speed following, asymmetric responses, delegated recommendations, and competitor-data commingling.

A command-line checker that scans generated documentation HTML for broken links, dead anchors, missing assets, duplicate IDs, and case-sensitive path mistakes.

An open-source command-line tool that opens websites in Chromium and audits SEO, performance, accessibility, security, and UX.

A public robotics experiment that trains a Unitree G1 humanoid in MuJoCo and runs its current simulation in the browser.

A browser-local toolkit for building, validating, comparing, and understanding CycloneDX cryptographic bills of materials.

An open-source memory method that turns notes, decisions, and corrections into searchable context for AI agents.

Ferromark turns Markdown into HTML for Rust and Node.js pipelines, while exposing metadata, rendering policies, and publishing-oriented features.

A bilingual simulation and minting system for designing neural circuits constrained by a versioned fruit-fly connectome.

ForgeRepo is presented as a free, self-hosted firewall and cache for npm, PyPI, Docker, NuGet, Maven, RubyGems, Composer, RPM, and APT.

Geodic connects iPhone and iPad users to MeshCentral-managed computers with remote desktop, terminal, file transfer, clipboard, and power controls.

A WordPress hardening and monitoring tool that checks files against official release hashes, records tamper-evident audit logs, and matches installed software against vulnerability advisories.

A software studio builds offline field apps, multilingual platforms, work-order systems, and cross-border tools for sectors where routine work has real consequences.

An open-source navigation layer lets Codex agents follow observed browser links and desktop controls toward a destination, then verify where they arrived.

A browsable showcase of 50 projects using Jev for games, creative experiments, automation, agents, data, and integrations.

A command-line tool for web search, open-source code search, library documentation, scraping, and crawling.

A build-to-order hardware site for running, fine-tuning, and training large AI models on controlled, local infrastructure.

A Python library that replaces personally identifiable information with stable placeholders before prompts reach an LLM, then restores the original values for tools and users.

Pollora is a PHP framework for building WordPress sites with Laravel-style routing, Blade templates, type-safe APIs, and automatic code discovery.

A local-first database IDE for querying, editing, comparing, and governing work across PostgreSQL, MySQL, SQL Server, MongoDB, Redis, and other engines.

An independent project site publishing Minecraft Java mods, small games, installation guides, and candid development notes.

A codebase-analysis tool that maps project structure, surfaces risky complexity, and compares findings across scans.

EleuScope is a local-first research workbench for moving from biological questions to inspectable analyses, figures, methods, and recorded decisions.

A macOS utility that keeps Claude Code, Codex, Antigravity, and other coding-agent sessions visible beside your working windows.

一家台中的小型團隊，替中小企業把 ERP、手機回報、對帳、告警與報表流程整合成部署在客戶自有主機上的營運系統。

An open-source macOS utility consolidates session, weekly, credit, and spend indicators from AI subscriptions, routing providers, and API keys.

A local macOS companion that lets command-line AI agents speak, ask questions, read text aloud, and translate it.

A practical developer reference and sandbox for testing modern email-validation rules.

Bender commissions egocentric RGB-D recordings of people performing real tasks, then delivers each accepted session as a single MCAP file.

A sourced guide separates Codex resets into context, configuration, authentication, and transport problems.

A device-aware sanitization and ITAD platform that records verification, certificates, quarantine events, and tamper-evident evidence for processed hardware.

A client-side toolbox for formatting JSON, testing regexes, decoding tokens, converting timestamps, and handling other everyday developer chores.

A dependency-mapping tool for small SaaS teams that translates vendor incidents into evidence-backed product-impact briefs for Slack.

MasterX combines ECU file analysis, map editing, automatic solutions, checksum correction, dyno tools, and a library of more than 246,000 original files.

RestoSignals turns liquor licenses, food permits, business filings, and other public signals into scored leads for restaurants and bars that may open in the next one to six months.

A searchable registry of humanoid and consumer robots with specifications, source records, verification states, comparisons, and a changelog.

ShardNet presents an open compute network for placing distributed training jobs across heterogeneous accelerator capacity.

A read-only API for public X profiles, tweets, threads, followers, and search, with agent-oriented formats and usage-based credits.

BoundID presents identity, delegation, spending controls, and execution guards for AI agents handling payments.

Bruno Bracaioli’s site publishes first-person work on AI agent security, MCP systems, and governed autonomy.

ChatRail connects WhatsApp numbers to APIs, webhooks, schedules, delivery logs, and replies grounded in the data attached to each alert.

A desktop and command-line utility for packing, verifying, signing, and protecting DayZ mod files.

An independent community firmware project that adds a rebuilt touchscreen interface, local-network printing, and a trusted-LAN Web UI to the UltiMaker 2+ Connect.

Flyscapee connects a modeled MaleCNS fruit-fly steering circuit to a shared 2004Scape client.

An open standard for proving that an authorized human approved an AI action before it affected the world.

A catalog of typed Choice, Score, and Noul schemas for routing model outputs in code.

A no-signup browser playground for testing typed Choice, Boolean, and Score decisions with the Jev AI model.

A JavaScript async primitive that combines a familiar promise API with laziness, cancellation, and a single-shot Observable model.

An observatory for how autonomous agents discover services, transact, and generate economic activity.

MarginCI runs the same test cases against two code revisions, measures supported paid API usage, and fails the check when cost per successful operation exceeds a configured threshold.

A local-first agent platform that plans, executes bounded actions, and records evidence in an append-only proof ledger.

A streaming-continuity service that fills signal gaps with clips and returns to live when the connection recovers.

An open distributed project that uses browser computation to generate, verify, store, and distribute a massive prime-number dataset.

An onchain trigger studio for Robinhood Chain that listens for events, tests conditions, and fires effects without requiring users to run a node.

VANTAR designs, integrates, and operates avionics, power, RF communications, flight software, and ground systems for small satellites.

A focused schema explorer that starts with one table, reveals connected relationships, lets you edit the design, and prepares migration scripts.

Quaqua Robotics records human interventions in robot fleets, groups them by underlying cause, and proposes feeding the resulting fixes back into the fleet’s existing software stack.

A Spanish-language phishing and email-fraud defense system that explains its verdicts, verifies doubtful requests, and automates containment.

A self-hosted MCP server that connects AI clients to tools for email, commerce, databases, knowledge storage, and WordPress.

A daily game presents seven logs from one hidden telemetry source and asks players to identify the suspicious one.