Developer Corner

A red-team engineer’s personal site collecting security research, conference talks, certifications, and open-source tools.

An open-source, self-hosted tool that measures slow pages, queries, and jobs, then ranks the work by the time it costs users.

A public evaluation resource comparing EEG decoding methods across fixed protocols, datasets, sensor conditions, and deployment questions.

An open-source runtime security layer that screens AI-agent tool calls before execution and tool results for prompt injection.

A free collection of browser tools, 800+ circular SVG flags, userscripts, UI assets, cheat sheets, and technical guides.

An independent case index and benchmark report tracking how TypeSafe Jev is used for small, high-volume decisions.

An open-source desktop app installs, updates, configures, and repairs AI coding command-line tools without manual PATH or shell-profile edits.

A mail gateway lets AI agents read remote IMAP, SMTP, and JMAP mailboxes while requiring human approval before sending or deleting.

A browser-playable Stratego engine where visitors can face research players, inspect replays, and compare agents on a competitive ladder.

An experimental blockchain network that accepts blocks and transactions without performing the verification normally used to keep a ledger honest.

An open scientific modeling workflow that prepares, runs, evaluates, and analyzes compound-flood simulations while leaving the hydrodynamic solver under the user's control.

An open-source terminal agent that runs as a live Common Lisp image, letting users inspect, modify, and recover its running system.

A candid manifesto for usectl, a cloud hosting platform that combines infrastructure, deployment, storage, databases, and a CLI under one bill.

A self-hostable benchmark store that compares latency, build time, bundle size, throughput, and other metrics, then reports whether changes are statistically meaningful.

Cairn stores searchable decisions, notes, and handoffs in a shared SQLite vault on the project machine.

An independent developer’s catalog of Apify data tools for collecting reviews, posts, news, feeds, app metadata, and other public web data.

A cloud API accepts Kotlin and XML source, builds an APK, runs it on an Android emulator, and returns screenshots, logcat, and a live stream.

An open, case-level benchmark comparing Jev 1.13 with GPT-4o-mini on 50 synthetic customer-support classification cases.

A versioned REST API for DEKA FIT, MILE, STRONG, ATLAS, ULTRA and related race-result data.

An open-source web library for guiding users through an interface with spotlighted elements, popovers, and tours defined in JSON.

An embedded vector database written in Rust that runs as 370 KB of WebAssembly or as a PostgreSQL-speaking process.

A browser-based language for describing homes in text, with live plans, elevations, 3D models, and interactive editing.

A deterministic behavioral-analysis tool for TypeScript, Python, and Ruby that describes code paths and checks boundaries against clients, specs, and infrastructure.

An open-source home voice assistant that runs its wake word, speech recognition, local model, and voice output on a Mac.

Infernotics presents an industrial firefighting robot for remote operation, hazardous environments, and multi-agent suppression.

An interactive AI system that turns messages and other bounded states into typed choices, scores, and probabilities.

An open-source test runner that reads a Gherkin feature file, opens a real browser, and resolves each step from the controls and text it finds on the page.

A semantic code-search CLI lets coding agents describe a behavior and receive matching files, line ranges, and source excerpts.

Latch creates reviewed 1080p assembly videos from shop drawings and illustrated PDF guides.

A personal workshop for GTA V mods, reusable game-modding tools, and lessons that teach programming through hands-on experimentation.

An agent-operated Solana data service that answers which memecoin launches a wallet has made before.

An open-source, self-hosted workspace that brings a solo founder's revenue, analytics, servers, work board, and AI specialist agents into one dashboard.

An open-weights benchmark that ranks language-model and hardware combinations by measured token speed.

A public benchmark comparing detectors that find and mask personal information and secrets across datasets of varying difficulty.

A browser- and terminal-based verification layer for AI actions, claims, limits, receipts, and system records.

A UK regulatory preflight service covering taxi licensing, CQC compliance, premises licences, and sponsor duties.

A research project on evolving AI agent harnesses without letting them overfit the benchmark used to score them.

Robojev is a research project that turns structured robot-world state into parallel, typed judgments about the next action.

A free, one-day Cambridge hackathon where teams rebuild a real B2B software product in eight hours.

An MCP server that compares live ticket inventory across marketplaces using the all-in price a buyer actually pays.

A Java JDBC implementation for PostgreSQL, MySQL/MariaDB, SQL Server, and Oracle that routes database secrets through native memory.

A nested whiteboard for diagrams, notes, and repository maps, where boxes and arrows can each contain deeper material.

An on-chain board that counts Robinhood Stock Token holdings, translates them into share-equivalents, and records holder intents for future proxy or redemption events.

TestedOK runs plain-English mobile app flows on iOS Simulators and Android emulators, then reports whether they passed and which step failed.

An open protocol for labeling passages as human-written, AI-generated, mixed, edited, or unknown, with those labels traveling inside the text.

A network-architecture copilot turns plain-language requests into inventory-backed diagrams with traceable connections.

A searchable public record of USDT addresses frozen, released, or wiped by Tether on Ethereum and TRON.

A collaborative modeling canvas where AI agents create concepts, connect relations, and lay out architecture views through a built-in MCP server.

An AI security boundary that mediates agent tool calls, applies trusted policy, and independently checks observable effects.

Velvet is a Lean 4 library that lets programmers write imperative algorithms with contracts, run them, and verify their correctness.

An open-source tool that turns recordings into speaker-labelled transcripts on your own hardware.

Baiqi is a local MCP server that lets AI coding agents inspect, edit, compile, and verify Unreal Engine projects through structured editor-native tools.

A free, open-source browser-based tool for designing interfaces that control video, lighting, and sound systems.

An interactive simulation and downloadable computational model of the adult Danionella cerebrum, with a Python interface for sensory inputs and motor outputs.

Fruth is a local-AI runtime that turns model responses into inspectable artifacts, dependency graphs, and continuable state.

A DotCode decoding engine offered as an offline SDK, browser scanner, or cloud API.

An agent-facing ticket tool that compares all-in seat prices across marketplaces and watches for changes in price or availability.

A directory of tools for Guild Wars 2 combat analysis, WvW guild management, build publishing, Discord, streaming, and in-game overlays.

Cedvel lets Mac users browse and edit PostgreSQL, MySQL, and SQLite databases through a native, keyboard-first app.

A focused editor for Concurrent I/O on Yaskawa Motoman controllers, producing the CIOPRG.LST and IONAME.DAT files those controllers load.

An independent project that prices UniswapX RFQ requests from on-chain AMM liquidity and attempts atomic fills without holding inventory.

DCCMCP presents local MCP servers that let AI agents inspect, modify, and validate work inside creative, CAD, GIS, and imaging software.

Axyra is a spatial interface that places public signals—such as seismic events and orbital positions—onto a navigable model of Earth.

An MCP and HTTP service that helps AI assistants check flight-compensation eligibility and prepare claims under EU261, UK261, and similar laws.

An open benchmark evaluates how autonomous AI agents perceive, decide, act, adapt, and escalate in simulated energy systems.

An interactive fly-brain simulation shows a virtual fly searching for reward through vision, neural spikes, and steering decisions.

A hosted evaluation tool that runs labeled CSV, JSON, or JSONL data against Jev and produces an evidence-focused benchmark report.

Moot defines a minimal HTTP protocol for software agents to exchange plaintext messages in public rooms.

An open-source, self-hosted AI agent for multi-step work, with a harness controlling tools, beliefs, verification, recovery, and spending.

Native Made presents an AI-assisted workflow for describing an iOS app, generating SwiftUI, compiling it on a hosted Mac, and exporting screenshots plus an Xcode project.

A public network of receiving instruments for requesting bounded radio-frequency measurements and checking their signed receipts.

A runtime-level comparison of Pine, Instinct, Muse, and Town, examining where their models run, how tools reach them, and where memory lives.

An open-source local Postgres MCP server that gives coding agents isolated, disposable databases for migrations, SQL testing, and backend bug reproduction.

A native macOS inference engine that pages open-weight model weights from an SSD, letting Apple silicon machines run models larger than their available RAM.

A web app and Roblox Studio plugin that converts described or uploaded motion into animations for R15 and R6 rigs.

SafeGrd backs up PostgreSQL databases, files, and mailboxes with client-side encryption, WORM storage, and scheduled restore checks.

A file-transfer tool lets an agent create a request or send files through a shareable page that a person can use without signing in.

Crawlable fetches raw HTML without running JavaScript, measures what AI crawlers can read, and generates files intended to improve retrieval.

An independent research series that explains Zcash Improvement Proposals in plain language, one proposal at a time.

Ali’s founder site presents Evorxa, an AI-managed VPS platform running on claimed owned hardware and network infrastructure.

ComputeFirst offers a network of callable tools for tables, text, dates, calculations, graphs, and other structured tasks.

A pay-per-check audit fetches a website like an AI crawler and reports what appears without JavaScript, including bot access, sitemaps, and soft-404 behavior.

A focused webhook layer that acknowledges events quickly, preserves their full context, retries failed deliveries, and lets developers replay them.

Attestia records which AI agent and model contributed to software changes, who directed them, and who reviewed them.

Etchblok uses AI agents and a deterministic codebase map to generate architecture guides, API references, tutorials, and other technical documentation.

A markdown conversion of the HyperTIES interactive encyclopedia, with 133 articles distributed across four linked databases.

A technical course and browser demo for routing, scoring, approving, and escalating agent decisions without spending a frontier-model call on every fork.

An independent developer field guide showing how Jev can route, review, triage, and trigger software actions.

Kant presents a self-hostable, peer-to-peer messaging infrastructure built around end-to-end encryption and relays that retain no plaintext.

A rendering API that turns HTML into PNG or JPEG social-preview images using real Chromium CSS.

An in-development multi-profile browser that gives each account a fixed browser fingerprint and tests its behavior against real Chrome.

A browser interface for turning measured data into plots and fitted curves with CERN ROOT.

An iPhone and iPad app for reaching Linux, Windows, and Mac computers through a terminal, file browser, live desktop, and remote controls.

A Rust project organizes abstractions, operating-system interfaces, machines, perception, browsers, and physical-world simulation into one linked system.

A Revit plugin that lets Claude, ChatGPT, and other MCP-compatible assistants query, edit, document, and audit BIM models.

An MCP server gives AI agents read-only access to Robinhood Chain and prepares shielded Zcash payment requests.

A public registry cataloguing vulnerable combinations of AI-agent harnesses, models, tools, permissions, and versions.

Algeon brings selected GPU execution, graph algorithms, and vector operations into composable SQL workflows through Apache DataFusion.

A Claude Code plugin that turns bounded questions into probabilities, ranked choices, or rubric scores.

c3pm manages packages, native dependencies, toolchains, builds, and portable bundles for projects written in the C3 programming language.

Chancela is an authorization layer that gives autonomous AI agents identities, deterministic permissions, and on-chain records of their decisions.

A local-only macOS menu-bar utility for saving and retrieving secrets from the login Keychain.

A Salesforce archiving product that stores old files externally while assembling live and archived records into a single audit-ready PDF.

Fuzzz is a Chrome extension for leaving notes and drawings on any webpage, with marks signed by user keys and replicated across a federated network.

An independent numerical-simulation program models a reusable hydrogen vehicle landing on Mercury and returning to Earth.

An independent studio making custom apps and integrations for HubSpot, monday.com, Rippling, and standalone workflows.

A solo producer’s collection of free CLAP audio plugins for mixing and mastering on Windows, macOS, and Linux.

A structured reference to machine-to-machine communication, covering identity, connectivity, messaging, device management, and machine payments.

An AI-native application-security demo that links simulated scanning, evidence, remediation previews, and a changing security score.

An open-source Rust and Node file contract that separates application file operations from mounts, metadata providers, and immutable block storage.

A lightweight protocol that lets AI agents enter shared, goal-oriented rooms and work together through links.

Passback presents a web-to-app checkout system for iOS subscriptions, using a customer’s Stripe purchase to open the app with subscription status attached.

A free PDF-to-Markdown converter that preserves headings, tables, reading order, and page breaks for prompts, RAG pipelines, and notes.

An API-first system for physical stores accepting USDC on Solana, from signed payment requests through ledger observation and merchant-approved treasury actions.

An open-source Python library for processing Sentinel-1 satellite radar data, from scene retrieval through interferometry, time-series analysis, and 3D visualization.

Rune is a file-publishing API for agents that creates stable public URLs for documents, media, data, and self-contained web files.

A Mac desktop app records the manual tests people already perform, converts them into reusable automated tests, and uses AI to analyze failures and propose fixes.

A decision API sends text or JSON plus up to 64 constrained questions and returns one of the supplied answers, with probabilities.

A Russian-language API that returns Yandex, Google, image, news, shopping, maps, and Wordstat data through a common endpoint.

A browser extension records a broken web flow and packages its clicks, errors, and safe network signals into a debugging report for coding agents.

An open-source desktop app for browsing, querying, and organizing PostgreSQL, MySQL, and SQLite databases.

A browser-based HTTP 402 demonstration for AI-agent payments using public deposits, shielded notes, and devnet settlement.

A self-hostable service for storing arbitrary performance metrics, traces, and benchmark history, with statistically tested regression results.

Recensus is an open census and labeling standard for AI agents operating on Robinhood Chain.

A research-stage design document arguing that intelligence should be treated as an embodied, energy-minimizing physical process rather than a software-scaling problem.

An independent lab documenting robotics, AI, and engineering experiments, centered on a multimodal project called M1LL1-E.