Trending archive for Nov 1, 2026

openai/math

OpenAI's model-generated math proofs, with Lean formalizations in progress — A collection of 719 mathematical manuscripts in 372 families produced by an internal OpenAI model tackling open research problems, released as model-development evaluations saturated on existing benchmarks. About 42% of top-line results have Lean formal proofs so far, with reasoning summaries and a comparator tool included for verification.

guregodevo/memdoor

A single-binary agent CLI that turns durable workflows into cheaper tokens — Memdoor is a Go-built CLI that runs coding/automation agents locally or over SSH with no Docker or Python needed. It stores multi-step jobs as repo files with checks, approval gates and resume, and uses a decision model to cut token costs. A published benchmark shows a 5-agent Memdoor workflow outscoring a single-call baseline on LiveCodeBench Hard.

kevinraymond/florilegium

Procedural Delft-blue flower margins drawn live inside LuaLaTeX PDFs — Florilegium generates Delft-blue floral page margins entirely in Lua at compile time—no images or imported paths, just parametric curves painted like watercolor. It offers both hand-authored compositions and a space-filling 'grown' layout that adapts to leftover page space, giving every document a unique garden.

herdrdev/herdr-app

A desktop terminal manager built around coding agents — Herdr organizes terminal work into spaces, tabs and panes, and surfaces which coding agents are Working, Blocked or Done across all your spaces in one sidebar. Terminals keep running after you close the window, and it pairs with a separate herdr-core component it can install for you. Built in Rust for Linux, macOS and Windows.

jepsen-io/particledb

Jepsen Puts ParticleDB's Transaction Consistency to the Test — A Jepsen test suite for ParticleDB, a Postgres-compatible database from Orkes. It currently focuses on single-node internal transaction consistency checks, useful for engineers evaluating ParticleDB's reliability claims before wider distributed testing is added.

nrslib/takt

YAML-defined workflows that discipline multi-agent coding loops — TAKT is a CLI that encodes planning, implementation, review and fix loops as YAML workflows, assigning each step its own role, context and transition rules rather than relying on one agent to remember everything. It runs tasks in isolated worktrees with traceable logs and human checkpoints, aimed at teams tired of babysitting AI coding agents like Claude Code or Codex.

aquasecurity/trivy

An all-in-one scanner for vulnerabilities, misconfigs, secrets and SBOMs — Trivy scans container images, filesystems, git repos, VM images and Kubernetes clusters for CVEs, IaC misconfigurations, exposed secrets, license issues and SBOM data in one tool. It plugs into GitHub Actions, a Kubernetes operator and a VS Code extension, making it useful for DevSecOps teams wanting a single scanner across their pipeline.

aquasecurity/trivy-action

Trivy Scanning Baked Into Your GitHub Actions Pipeline — This official GitHub Action wraps Trivy to scan container images, filesystems, Git repos, IaC configs, and private registries, and can also generate SBOMs. It integrates directly with GitHub Code Scanning, making it a quick way to add vulnerability checks and fail CI builds on findings without installing Trivy manually.

web-push-libs/web-push

Node.js Toolkit for Standards-Based Web Push Notifications — This library implements the Web Push Protocol and its message encryption spec, letting a Node.js backend send push notifications with VAPID keys and encrypted payloads, plus legacy GCM support. It includes a CLI for generating VAPID keys and sending test notifications without writing code.

yuiseki/map-from-scratch

A from-scratch skill that lets AI agents build one-off, minimal map renderers — Instead of one general map engine, this gives AI agents a skill and harness to rebuild the smallest, fastest renderer for each specific request, picking language/API/runtime per case (Canvas2D, WebGL2, C+EGL, even microcontrollers). Each example is defined in spec.yaml with measurable requirements and benchmarks that the harness checks pass/fail, drawing on techniques from 48 reference map engines.