temporalio/sdk-python 1.34.0 — Temporal SDK for Python
A look at the latest release of temporalio/sdk-python and the role of the SDK for writing Temporal workflows and activities in Python.
CURATED OPEN SOURCE INTELLIGENCE
A look at the latest release of temporalio/sdk-python and the role of the SDK for writing Temporal workflows and activities in Python.
spacering-net/codeg is a workspace that handles the conversations, file changes, and sessions of multiple AI coding agents in one place.
GenieX in qualcomm/geniex is a generative AI inference runtime that runs LLMs and VLMs locally on Snapdragon devices.
justrach/nanoterm is a Zig project whose block-based terminal input field handles both shell commands and ACP-based coding agent requests.
b0yko/tern is a local tool for Apple Silicon that searches podcast and video libraries by speech, on-screen text, and scenes, then lets you edit and export the results.
Introduces the public release status of smtg-ai/claude-squad v1.0.20 and how to manage multiple local coding agents from the terminal.
openai/openai-agents-js is a framework for building multi-agent workflows in JavaScript and TypeScript.
A look at the deep learning operations covered by pytorch/fbgemm's three packages and at what cannot be confirmed from the available materials.
An introduction to the structure in which cv-superding/hypercontext-mode handles tool-output sandboxing, conversation compression, and persistent memory within a single execution environment, plus the search measurements in its README.
Summarizes the browser-based SVG editing features and structure of svg-edit/svgedit, and notes that the changelog for v.7.3.3 was not provided.
ggml-org/llama.cpp b11750 is a release that fixed compile warnings in the SYCL code.
A self-hosted agent memory service presented by the hsuanshao/agentmemory repository, built around L0–L3 layered memory and hybrid retrieval.
Covers what qwenaudio/sensevoice's public SenseVoiceSmall model and the latest release runtime-llamacpp-v0.2.1 do, and the limits that can actually be verified.
ExecuTorch is an open-source stack for running PyTorch models locally on mobile, edge, and embedded devices.
A repository that lets multiple AI coding agents check the same project rules through an AGENTS.md contract template, a dependency-free Python linter, and a session start hook.
qwenlm/qwen-agent is a Python framework for building LLM applications based on Qwen's tool use and planning capabilities.
textlint/textlint is a plugin-style natural language linter that installs rules separately to check Markdown and plain text, and its latest release is v15.8.0.
rapidai/rapidocr is a Python toolkit that helps deploy OCR models across multiple execution engines and devices.
Supacode is a macOS app that runs multiple coding agents in parallel, each in its own worktree and terminal.
steel-dev/steel-browser is an open-source browser API that helps AI apps and agents manage sessions and automate the web.
No matching stories.