TesterArmy e2e adds per-attempt branches to Chromium microVMs

#1 🐙 tester-army/e2e

tester-army/e2e - tester-army/e2e is TesterArmy’s Apache-2.0 framework for web and mobile testing, pairing natural-language agent actions with locators and assertions; the repo includes its SDK, engines, docs, examples, and reusable skills. On October 9, 2026, tester-army released @e2e-dev/smol@0.1.0, contributed by BinSquare, adding local Chromium microVMs with one warm browser per worker and default per-attempt branches created in about a second. Worth trying if you need isolated, agent-assisted end-to-end tests. Release notes

#2 🐙 cursor/plugins

cursor/plugins - **cursor/plugins** is Cursor’s repository of official plugins for developer tools, frameworks, and SaaS products, each with its own manifest. It covers plugin scaffolding, reviews, compatibility scans, parallel agent orchestration, SDK automation, and service integrations; the repository has 10,996 stars and a 1,140 weekly trending gain. Worth trying if you need packaged Cursor workflows or integrations.

#3 🐙 mksglu/context-mode

mksglu/context-mode - **mksglu/context-mode** is an MCP server that limits raw tool output in AI coding-agent context windows and preserves relevant session memory. It includes plugins, hooks, configuration, documentation, and 11 MCP tools for sandboxing output, search, diagnostics, and maintenance across 17 supported clients. Worth trying if you need to manage context use and continuity in coding agents.

#4 🐙 open-webui/open-webui

open-webui/open-webui - **open-webui/open-webui** is source for a self-hosted AI app that connects to Ollama and OpenAI-compatible APIs, with tools for agents, RAG, voice, image generation, analytics, and access control. On October 10, 2026, open-webui released v0.12.0, adding OpenAI Realtime voice calls, animated 3D avatars, and authenticator-app two-factor sign-in. Worth trying if you need a provider-agnostic interface for local and cloud models. Release notes

#5 🐙 pbakaus/impeccable

pbakaus/impeccable - pbakaus/impeccable is design guidance for AI coding agents, containing 1 skill, 24 commands, live browser iteration, and 59 deterministic detector rules that can run without an LLM or API key; the repository has 80,115 stars. pbakaus released Impeccable skill-v4.5.2 on 2026-10-09 at 15:30:51Z, adding code-first app screens and documents, HTML skeleton sketches, and Live Mode for Astro 7.3.6 and later. Worth trying if you need reusable guidance and checks for AI-generated frontend design. Release notes

#6 🐙 mattpocock/skills

mattpocock/skills - **mattpocock/skills** is an MIT-licensed collection created by mattpocock of small, adaptable, composable agent skills for software engineering, with setup instructions for Claude Code, Codex, GitHub Copilot, Gemini CLI, and other agents. It includes reusable workflows for requirements grilling, shared documentation, test-driven development, and bug diagnosis. Worth trying if you need model-compatible building blocks for structured coding-agent workflows.

#7 🐙 mvschwarz/openrig

mvschwarz/openrig - mvschwarz/openrig is an Apache-2.0 system for defining persistent Claude Code and Codex agent teams in YAML, with CLI packages, docs, a demo, tests, and skills. mvschwarz released version 0.6.10 on October 12, 2026, as a 56-change fix release, including 38 changes from ten outside contributors. Worth trying if you need coordinated coding-agent teams and can provide Node.js 22 or 24 and tmux on macOS or Linux. Release notes

#8 🐙 heygen-com/hyperframes

heygen-com/hyperframes - **heygen-com/hyperframes** is an Apache-2.0 open-source framework for turning HTML, CSS, media, and seekable animations into deterministic MP4 videos, with a CLI and 21 published skills for AI coding agents. heygen-com released HyperFrames v0.8.149 on 2026-10-11, adding importable `transcribe` and `removeBackground` functions that return typed results and throw typed errors; the repo has 61,440 GitHub stars and 4,223 weekly trending gains. Worth trying if you need agent-assisted, code-based video rendering. Release notes

#9 🐙 michaelshimeles/skills

michaelshimeles/skills - michaelshimeles/skills, shared by michaelshimeles, offers seven reusable agent skills and an AGENTS.md workflow template for Claude Code, Cursor, and Codex. It covers isolated worktrees, service-layer design, evidence capture, before/after comparisons, Greptile review loops, and prose cleanup. Worth trying if you need a structured isolate-build-prove-ship workflow for coding agents.

Also covered by: @Peter Yang

#10 🐙 headroomlabs-ai/headroom

headroomlabs-ai/headroom - headroomlabs-ai/headroom compresses tool outputs, logs, RAG chunks, files, and conversation history locally before they reach an LLM. The Apache-2.0 repository includes Python and TypeScript libraries, a local proxy, agent wrappers, an MCP server, and local retrieval of original content. Worth trying if you need to reduce agent context without sending data elsewhere for compression.

#11 🐙 promptfoo/promptfoo

promptfoo/promptfoo - promptfoo/promptfoo is an open-source, MIT-licensed CLI and library for evaluating prompts, models, agents, and RAG applications, with red teaming and vulnerability scanning. It includes automated comparisons, CI/CD checks, pull-request code scanning, and examples, while running evaluations locally. Worth trying if you need repeatable LLM testing and security checks in development workflows.

#12 🐙 langfuse/langfuse

langfuse/langfuse - langfuse/langfuse is an open-source platform for tracing, monitoring, evaluating, and debugging LLM applications. It includes prompt management, datasets, an LLM playground, APIs, Python and JavaScript/TypeScript SDKs, integrations, and cloud or self-hosted deployment options. Worth trying if you need one system for LLM observability and evaluation.

#13 🐙 mlflow/mlflow

mlflow/mlflow - mlflow/mlflow is an Apache-2.0 open-source platform for tracing, evaluating, monitoring, and optimizing agents, LLM apps, and ML models. It includes prompt management, 50+ evaluation metrics and LLM judges, an AI Gateway, experiment tracking, a model registry, deployment tools, and tracing integrations for more than 60 frameworks. Worth trying if you need one system for AI observability and model lifecycle management.

#14 🐙 n8n-io/n8n

n8n-io/n8n - n8n-io/n8n contains source for a fair-code platform for building AI agents and workflows using a visual canvas, JavaScript, Python, and multiple model providers, with self-hosted or cloud deployment. n8n-io released n8n@2.42.7 on October 12, 2026, fixing concurrent credential-test isolation and resumption of tool calls for deferred tools loaded by a skill. Consider it if you need extensible AI workflow automation without model lock-in. Release notes

#15 🐙 firecrawl/firecrawl

firecrawl/firecrawl - **firecrawl/firecrawl** is an open-source web-data toolkit and hosted service for AI agents, offering search, scraping, crawling, browser interaction, and structured outputs. The AGPL-3.0 repository includes apps, a CLI, reusable skills, workflows, and examples; worth trying if your AI product needs web content or page-interaction capabilities.

#16 🐙 ollama/ollama

ollama/ollama - ollama/ollama is an MIT-licensed project for running and managing open models through an app, CLI, REST API, and Python and JavaScript libraries, with source and documentation for macOS, Windows, Linux, and Docker. Ollama released Ollama v0.40.3 on 2026-10-11 at 01:39:04Z, removing Claude connector and Codex service-tier warnings and pausing a background model upgrade related to embedding and other request performance. Worth trying if you need a local model runner with API and coding-agent integrations. Release notes

#17 🐙 microsoft/BitNet

microsoft/BitNet - microsoft/BitNet is Microsoft’s MIT-licensed bitnet.cpp framework for lossless 1.58-bit LLM inference on CPUs and GPUs. It includes source and GPU code, optimized kernels, documentation, setup tools, and inference and server scripts. Worth trying if you need local inference for supported BitNet models.

#18 🐙 NVIDIA/garak

NVIDIA/garak - NVIDIA/garak is an Apache-2.0 command-line vulnerability scanner that uses static, dynamic, and adaptive probes to test LLMs and dialog systems for prompt injection, data leakage, hallucination, toxicity, misinformation, and jailbreaks. The repository includes Python source, signatures, documentation, tests, tools, examples, and reporting code, with support for Hugging Face, Replicate, OpenAI API, AWS Bedrock, LiteLLM, and REST interfaces. Worth trying if you need red-team checks across multiple model interfaces.

#19 🐙 thedaviddias/Front-End-Checklist

thedaviddias/Front-End-Checklist - **thedaviddias/Front-End-Checklist** is an MIT-licensed front-end quality system with 386 English rules across 11 active categories, available through a website, README, and hosted server with 11 MCP tools. Its MCP integration supports code review, live-URL audits, rule lookup, and remediation workflows, while the repository also provides reusable skills for broad audits and focused rule-specific guidance. Worth trying if you need structured frontend checks for human or AI-agent workflows.

#20 🐙 Graphify-Labs/graphify

Graphify-Labs/graphify - Graphify-Labs/graphify is an Apache-2.0 CLI and reusable assistant skill that turns code, docs, SQL schemas, configs, and PDFs into a queryable knowledge graph using local, deterministic code parsing rather than a vector store. Graphify-Labs released Graphify v0.9.86 on October 11, 2026, with 18 community pull requests covering extraction fixes, safety and determinism hardening, and CLI/query usability. Worth trying if you need explainable project maps for AI coding assistants. Release notes