The Mog Programming Language
First language designed for LLMs to modify safely with capability permissions.
Zero to OneWizardryBig Brain
belisarius222
163833mo ago
Developer Tools Projects
Developer tools and utilities from Show HN — CLIs, editors, debugging tools, build systems, and developer productivity.
GitHub
A CSS 3D engine for the DOM. Renders polygon meshes in HTML by leveraging matrix3d transforms.
614TypeScript
●●●●Gem
A CSS 3D Engine (no WebGL)
CSS-only 3D engine that renders OBJ files—same constraint-craft energy as X86CSS.
WizardryZero to OneRabbit Hole
rofko
802714d ago
GitHub
Battery info program for x86-64 Linux laptops in 307 bytes of machine code
100Assembly
●●●●Gem
A (marginally) useful x86-64 ELF executable in 298 bytes
298 bytes total — direct syscalls, no libc, hand-written assembly that actually works.
WizardryZero to OneNiche Gem
meribold
1302mo ago
An application stack Claude coded directly in LLVM IR
Writing working apps in LLVM IR for AI agents skips human-centric abstractions entirely.
Zero to OneWizardryBig Brain
dboreham
1003mo ago
GitHub
Tiny HTTP-only file fetcher without libc, DNS or dependencies. Linux + Android. ~2KB static binary. One job, done small.
6C
●●●●Gem
I wrote a ~2KB executable file HTTP file downloader without Libc
Fitting an HTTP client into 2KB without libc is demoscene-grade constraint craft.
WizardryBig Brain
DenisDolya
962mo ago
One grammar, 18 YAML parsers – a Futamura projector in Common Lisp
One grammar generates 18 spec-compliant YAML parsers via Futamura projection in Common Lisp.
WizardryBig BrainZero to One
johnagrillo62
413mo ago
GitHub
Minimalist AI assistant in 6,832 bytes of x86-64 assembly. No libc, no runtime, no allocator
15Assembly
●●●●Gem
PlanckClaw an AI agent in 6832 bytes of x86-64 assembly
6KB binary for an AI agent—fits on a floppy disk 62 times over.
WizardryBig BrainZero to One
frntn
442mo ago
C4 – Separate file description from file content
Email a 2MB text file describing an 8TB project — then cd into it.
Zero to OneBig BrainWizardry
joshuak
412mo ago
GitHub
GPU-Native Programming Language. 4.5 MB binary. Any GPU. Zero dependencies.
5
●●●●Gem
OctoFlow – A GPU-native programming language
GPU-first language where CPU handles only I/O, shipping as single 2.2MB binary.
Zero to OneWizardryBig Brain
mr_octopus
343mo ago
Whistler – a Lisp that compiles to eBPF
Replaces LLVM for eBPF with a 5,500-line Common Lisp compiler matching clang optimization.
WizardryZero to OneBig Brain
atgreen
302mo ago
GitHub
Five SQLite-compatible engines (C/Rust/Zig/Go/Python + WASM) from one language-neutral spec — a LEAP research artifact
15PLpgSQL
●●●●Gem
I built five SQLite engines from one spec in a week
Five SQLite engines from one spec, all passing 99% of the upstream sqllogictest corpus.
WizardryBig BrainZero to One
safitudo
301mo ago
Relic, a tiny coding agent for ancient and constrained systems
Claude Code-style agent running on Windows 95 with 4 MB RAM is genuinely wild.
WizardryZero to OneRabbit Hole
felixrieseberg
319d ago
GitHub
Admiran is a pure, lazy, functional language and self-hosting compiler based upon Miranda
111Assembly
●●●●Gem
Admiran: a pure, lazy functional language and self-hosting compiler
Self-hosting Miranda dialect compiler in 7kloc with 19-pass optimization and GC.
WizardryZero to OneBig Brain
taolson
203mo ago
GitHub
the MOS tech 6502 8-bit microprocessor in pure SQL powered by postgres
63PLpgSQL
●●●●Gem
MOS Tech 6502 emulated in pure Postgres
Pure SQL 6502 emulation with opcodes as stored procedures—no external code anywhere.
WizardryZero to OneRabbit Hole
heywinit
202mo ago
WebAssembly Interpreter in a Header
Single-header WebAssembly 3.0 interpreter beats Wasmtime on embed complexity.
WizardryNiche GemZero to One
lifthrasiir
201mo ago
GitHub
Parse partial JSON as it streams in. Type-safe, tiny, built for LLM output.
3TypeScript
●●●●Gem
Jsonchunk – Parse incomplete JSON from streaming LLM responses
Missing primitive: tolerant JSON parser for streaming LLM output, typed and <1KB.
Zero to OneSolve My ProblemCozy
jbingen
103mo ago
Text Post
I’ve spent the last few weeks pushing the limits of .NET 10 (Native AOT) to see if I could achieve deterministic, hard real-time performance for large-scale parameter indexing. Today, I hit the 1,000,
●●●●Gem
Nexus – A 1KB .NET 10 core with <15μs latency for 1M parameters
1KB DLL hitting 12.8μs on 1M entries with live telemetry proving it.
WizardryBig BrainNiche Gem
NexusCore
103mo ago
Building a web server in assembly to give my life (a lack of) meaning
Hand-rolled ARM64 assembly server with range requests and slowloris mitigation.
WizardryBig Brain
imtomt
4302271mo ago
Git bayesect – Bayesian Git bisection for non-deterministic bugs
Bayesian git bisect handles flaky tests where standard bisect fails completely.
Big BrainNiche Gem
hauntsaninja
337432mo ago
I made a Clojure-like language in Go, boots in 7ms
Runs a Clojure dialect in 10MB binaries while Babashka needs a 30x larger JDK.
WizardryBig BrainNiche Gem
marcingas
292851mo ago
Free Alternative to Wispr Flow, Superwhisper, and Monologue
Replaces three paid apps with free Groq-powered dictation, context-aware name spelling included.
Ship ItSolve My Problem
zachlatta
2771323mo ago
I ported Tree-sitter to Go
90x faster incremental parsing in pure Go—tree-sitter without C toolchain friction.
WizardryBig Brain
odvcencio
2221063mo ago
GitHub
Ollama for classical ML models. AOT compiler that turns XGBoost, LightGBM, scikit-learn, CatBoost & ONNX models into native C99 inference code. One command to load, one command to serve. 336x faster than Python inference.
684Python
●●●Banger
Timber – Ollama for classical ML models, 336x faster than Python
336× faster tree model inference; compiles sklearn/XGBoost to C99, serves like Ollama.
WizardrySolve My Problem
kossisoroyce
207333mo ago
GitHub
Emdash is the Open-Source Agentic Development Environment (🧡 YC W26). Run multiple coding agents in parallel. Use any provider.
4,751TypeScript
●●●Banger
Emdash – Open-source agentic development environment
Parallel agents in git worktrees over SSH. Claude Code meets Tmux.
Ship ItWizardry
onecommit
206713mo ago
Tilde.run – Agent Sandbox with a Transactional, Versioned Filesystem
Roll back any agent run like a Git commit with a single command.
SlickSolve My ProblemBig Brain
ozkatz
2051331mo ago
GitHub
enject: Hide .env secrets from prAIng eyes: secrets live in local encrypted stores (per project) and are injected directly into apps at runtime, never touching disk as plaintext.
498Rust
●●●Banger
enveil – hide your .env secrets from prAIng eyes
Stops AI tools from reading .env files by never storing secrets as plaintext on disk.
Solve My ProblemNiche GemShip It
parkaboy
2011313mo ago
RetroTick – Run classic Windows EXEs in the browser
Full Windows 95 emulation in the browser beats bare nostalgia.
WizardryEye CandyRabbit Hole
lqs_
195623mo ago
Trust – Coding Rust like it's 1989
Full Rust IDE running in a blue-screen TUI that can actually build itself.
CozyWizardry
wojtczyk
178901mo ago
ShadowCat – file transfer through QR Codes in a Browser
Streaming files over QR codes with missing-chunk recovery in a single offline HTML file.
WizardryNiche GemCozy
unprovable
1686524d ago
µJS, a 5KB alternative to Htmx and Turbo with zero dependencies
5KB drop-in AJAX library beats HTMX (16KB) and Turbo (25KB) with zero dependencies.
SlickBig Brain
amaury_bouchard
161923mo ago
ProofShot – Give AI coding agents eyes to verify the UI they build
Finally, visual proof when your AI agent claims it finished the UI work.
Solve My ProblemZero to OneSlick
jberthom
1611062mo ago
GitHub
NUMA-Aware Contention-Free Dynamically-Auto-Tuning Bash-Native Streaming Parallelization Engine
356Shell
●●●Banger
Forkrun – NUMA-aware shell parallelizer (50×–400× faster than parallel)
NUMA-aware bash parallelizer hits 400× GNU Parallel speeds with verifiable builds.
WizardryBig Brain
jkool702
151412mo ago
GitHub
Turn any MCP, OpenAPI, or GraphQL server into a CLI — at runtime, with zero codegen
2,226Python
●●●Banger
Mcp2cli – One CLI for every API, 96-99% fewer tokens than native MCP
MCP-as-CLI cuts tool schema tokens 96-99%, discovered on demand not injected.
Big BrainSolve My ProblemZero to One
knowsuchagency
1461003mo ago
We analyzed 1,573 Claude Code sessions to see how AI agents work
First analytics layer for Claude Code revealing 26% session abandonment rate.
Zero to OneDark Horse
keks0r
144863mo ago
GitHub
Mathematical animations for the web — Manim in TypeScript
392TypeScript
●●●Banger
I ported Manim to TypeScript (run 3b1B math animations in the browser)
Manim in TypeScript—zero-setup math animations at 60fps, API-compatible with Python.
WizardryZero to OneSolve My Problem
maloyan
140253mo ago
Skill that lets Claude Code/Codex spin up VMs and GPUs
Gives Claude its own machines—VM primitive for AI agents beats hogging your ports and RAM.
Solve My ProblemWizardryNiche Gem
austinwang115
138364mo ago
Unfucked (unf) – a filesystem flight recorder that snapshots every save
Continuous filesystem snapshots with AI burst detection—Git for unsaved work.
Solve My ProblemWizardryShip It
cyrusradfar
137913mo ago
GhostBox – disposable little machines from the Global Free Tier.
Aggregates spare global compute into a single CLI for disposable agent workspaces.
Zero to OneBold BetSolve My Problem
keepamovin
1261401mo ago
Ableton Live MCP
Arbitrary Python eval inside Ableton beats rigid predefined tool sets for real creative control.
WizardryNiche GemBig Brain
bschoepke
123791mo ago
GitHub
Portable DNS resolver in Rust — .numa local domains, ad blocking, developer overrides
1,293Rust
●●●Banger
I built a DNS resolver from scratch in Rust – no DNS libraries
Auto-TLS for local domains beats mkcert and /etc/hosts without manual cert management.
WizardrySolve My ProblemDark Horse
rdme
115682mo ago
GitHub
Write your BPF programs in Go, not C. gobee transpiles a Go subset to BPF C and generates typed cilium/ebpf bindings.
328Go
●●●Banger
Write your BPF programs in Go, not C
Go-to-BPF transpilation beats Aya's Rust approach by reusing clang's mature backend.
Big BrainSolve My Problem
boratanrikulu
1125224d ago
Agent framework that generates its own topology and evolves at runtime
Rejects GCU theater; headless async agents for real 9-to-5 work like reconciliation.
Big BrainBold BetSolve My Problem
vincentjiang
107354mo ago
GitHub
Open-source WYSIWYG .docx editor library with canonical OOXML, tracked changes, and real-time collaboration. Agent-ready.
1,888TypeScript
●●●Banger
Open-source .docx editor library for building document apps
Canonical OOXML parsing beats HTML conversion by preserving document semantics and layout fidelity.
SlickSolve My ProblemDark Horse
thisisjedr
1071724d ago
Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud
Fork-based VM isolation for parallel AI agents beats fighting over localhost resources.
Big BrainSolve My ProblemBold Bet
nab
1047910d ago
Nibble
Single-pass LLVM frontend in 3000 lines of C with zero heap allocations.
WizardryBig BrainNiche Gem
glouwbug
102251mo ago
Pion/handoff – Move WebRTC out of browser and into Go
Mocks browser WebRTC API to hand off sessions to Go for server-side recording or injection.
WizardryBig Brain
Sean-Der
101172mo ago
GitHub
Localhost service isolation and orchestration for git worktrees.
401Rust
●●●Banger
Coasts – Containerized Hosts for Agents
Git worktree + docker isolation solves the AI agent localhost port collision nightmare.
Big BrainSolve My ProblemZero to One
jsunderland323
99382mo ago
Agentic interface for mainframes and COBOL
Brings AI agents to TN3270 terminals to decode JES abends and write column-strict JCL automatically.
Zero to OneSolve My Problem
sai18
97511mo ago