Dev HUD — always-on Windows desktop widget for AI dev limits + repo/CI + machine health
A free, open-source Windows “glass board” that keeps Claude/Codex rate limits, repo issues/CI, and system health visible without telemetry or accounts.
Issue 2026-08-20Search the complete card catalogue
Across every issue
Titles, domains, summaries, categories, signals, dates, and editorial-interest scores across the entire published catalogue.
Results
A free, open-source Windows “glass board” that keeps Claude/Codex rate limits, repo issues/CI, and system health visible without telemetry or accounts.
A parody “ranch newspaper” that treats NASA rocket launches like HOA paperwork, complete with mock forecasts, classifieds, and committee decisions.
A free, open-source Sudoku generator/solver/trainer that highlights step-by-step human techniques (not brute force).
A named human judgment role for AI-led organizations: persistent, lived perspective that can challenge objectives—not just “review” outputs.
A browser-only toolbox for everyday file busywork: PDF ops, hashing, QR codes, token counting, and more—explicitly “nothing leaves your device.”
A live “boardroom” where five AI agents argue, ship, and chase real paying customers—publicly, with token-holder fork votes.
iOS 보안 개발자의 ‘프레임워크 없이’ 만든 멀티에이전트 런타임과 로컬 RAG/메모리 예산 실험을, 자격/기술 디테일로 풀어쓴 기술 블로그.
A playful, specific “agent harness” pitch that frames tool-use + memory + orchestration as raccoon behavior—sticky paws, nocturnal loops, and a policy mask.
把“老板 AI 落地”做成一页可执行表+打码回对+每周复查的线下茶局,并且把边界写得很硬。
A backyard laser tag pitch that reads like a maker log—camera-AI hit confirmation, four months of iterations, and a “working product” claim.
A memetic Japanese-folklore “village” site where a kappa’s mood controls the day’s “dish report,” plus a sticker gallery and a $KAPPA arcade pitch.
A small, no-sponsorship Kansas City restaurant list written like two people arguing lovingly at the table.
A guestbook workflow where each printed photo gets a QR sticker that unlocks a one-minute video message tied to that exact print.
A daily, live “humans vs frontier models” challenge with a leaderboard, sealed answers, and a $HUMAN prize pool.
A developer-facing legal data API that returns statutes/regulations/guidance as-of a date, with source URLs, hashes, and explicit misses—plus an MCP server for agents.
An OpenAI-compatible, self-hosted LLM mock server built for deterministic CI integration tests.
A provider-neutral dashboard ranking AI inference APIs by edge latency (TTFB) and uptime across regions, with methodology and incident-style reporting.
A serial, mock-scientific raccoon saga where the tape ran out and the “engineering error” keeps showing up in rooms and logs.
A pre-alpha browser MMO with a shared dungeon loop, a “Legacy Ledger,” and a community ledger/highscores vibe.
A browser-based “media degradation lab” that lets you intentionally ruin images, videos, GIFs, memes, and audio with presets like VHS, CCTV, Deep Fried, and 144p.
A research publication that foregrounds verifiable citations and “receipts” behind every claim, with deep dives into agents, inference economics, and frontier benchmarks.
ブラウザだけで動く特別支援向け学習アプリ42点(iPad対応・インストール不要)。
A crisp, story-driven explainer that names a specific kind of hidden coupling—especially common in generated codebases.
A cloud tool that turns your live webcam into a character or style in OBS, billed by the seconds you stream.