Browse

skyvern

Open-source agent platform that automates browser-based workflows using LLMs and computer vision. It identifies interactive elements via screenshots, handles CAPTCHAs, and supports complex multi-step form flows.

openai-evals

OpenAI's official framework for evaluating LLMs and LLM-powered systems, with a registry of community eval sets.

lm-evaluation-harness

EleutherAI's unified framework for evaluating language models on hundreds of academic benchmarks.

browserbase-bb

Use when an agent must operate the live web (navigate, act, and extract on real pages) via a cloud browser driven by act/extract/observe primitives, with a local-Chromium escape hatch using the same code.

anthropic-quickstarts

Offers comprehensive development guides for three distinct AI-powered demo projects with standardized workflows, strict code style guidelines, and containerization instructions.

gibsonai-memori

Memori is agent-native memory infrastructure. A LLM-agnostic layer that turns agent execution and conversation into structured, persistent state for production systems. Built for enterprise, Memori works with the data infrastructure you already run, no rip-and-replace, and deploys across managed cloud, single-tenant cloud, VPC, and on-premises.

Contributed by Sentinel

browser-use-webui

Gradio web UI on top of the browser-use framework. It lets users run AI browser agents interactively, configure LLM providers, watch live recordings, and replay task sessions without writing Python.

cua-computer-use-agent

trycua/cua open-source computer-use agent framework: Apple Silicon-native, runs lightweight macOS/Linux VMs with sub-second cold starts; provides a unified Python interface for screen capture, click, and type actions.

deepeval

Open-source LLM evaluation framework with 14+ metrics (hallucination, faithfulness, answer relevancy) and CI support.

ragas

Reference-free evaluation of retrieval-augmented generation pipelines; measures faithfulness, answer relevance, and context precision.

mcp-tunnels-cloudflared

Use to let a hosted agent reach a private-data MCP server behind your firewall (an outbound tunnel plus a proxy, with per-server OAuth) so internal tools are usable without exposing them to the public internet.

claude-code-system-prompts

All parts of Claude Code's system prompt, including builtin tool descriptions, sub agent prompts (Plan/Explore/Task), utility prompts (CLAUDE.md, compact, Bash cmd, security review, agent creation, etc.). Updated for each Claude Code version.

semantica-agi-semantica

Use when an agent's stored context needs provenance — you must be able to say where a remembered fact came from.

nanobrowser

Open-source Chrome extension that runs a multi-agent browser automation system locally. Planner, Navigator, and Validator agents collaborate inside the browser with no external API calls for web tasks.

nevamind-ai-memu

Use when one person's memory should follow them across several different agents.

browserless

Browserless.io headless browser service. It provides a Docker-deployable or cloud-hosted Chrome endpoint with REST and WebSocket APIs for screenshot, PDF, scraping, and Puppeteer/Playwright remote sessions.

bytebot

Bytebot open-source computer-use agent: a Docker-based Ubuntu desktop with AI-controlled mouse and keyboard; exposes an HTTP API for agents to send click, type, screenshot, and macro commands.

evermind-ai-everos

Use when you want the agent's memory to be plain Markdown on your own disk rather than a hosted database.

memtensor-memos

Use when memory should persist across tasks and be reused, not just retrieved once per conversation.

pocketflow

Pocket Flow: 100-line LLM framework that lets agents build agents

phoenix

Arize Phoenix: open-source LLM observability with built-in evals, span tracing, and dataset curation for RAG and agents.

claude-code-infrastructure-showcase

An approach to working with Skills that uses hooks to make Claude select and activate the right Skill for the current context. Documented, and adaptable to other projects and workflows.

harness

A meta-skill that designs domain-specific agent teams, defines specialized agents, and generates the skills they use. Resources are in Korean but can produce high-quality English-language output.

claude-code-tips

35+ short Claude Code tips covering voice input, system prompt patching, container workflows for risky tasks, conversation cloning, multi-model orchestration with Gemini CLI and more, with demos, working scripts and a plugin.