AIエージェント
qwen-audio-agent
Real-time voice runtime that keeps AI agents talking, working, and present.
★ 2.8k スターApache-2.0
ガイドを見る →agent-vaultは、Infisicalが公開しているオープンソースのAIエージェントプロジェクトです（GitHub: Infisical/agent-vault）。GitHubスターは2,323、フォークは163で、NOASSERTIONライセンスの下で公開されています。GitHubでの最終更新は2026年10月4日です。
Infisical's credential vault and HTTP proxy, so AI agents call APIs without ever holding your real secrets.
Decide which credentials your agents should be able to use indirectly, and where the vault will run, ideally on a separate machine from the agents.
Install Agent Vault with the official install script or the Docker image, and start the server with a master password.
Create a vault, add credentials and service rules that substitute dummy values with real ones, then create an agent and its token.
Set the agent environment variables, install the CLI on the agent side, and run your agent through Agent Vault so its requests are brokered.
Agents hold dummy placeholder values; the proxy substitutes real credentials into headers on outbound requests to services such as LLM providers and GitHub.
Use the built-in local encrypted store, or back a vault with an external secrets store such as Infisical, including dynamic secrets.
Works with existing MCP tools, CLIs, SDKs, and APIs by bootstrapping the agent environment with HTTPS_PROXY, with no code changes to the agent.
Control which services and endpoints each agent may reach, and optionally deny any request that matches no service rule.
Authenticated traffic flowing through the proxy can be inspected to monitor and diagnose agent behaviour.
The server exposes a management API/UI on port 14321 and the MITM proxy on 14322, and a TypeScript SDK lets orchestrators mint short-lived tokens for sandboxed agents.
Most deployments use the install script or Docker image rather than a source build.
git clone https://github.com/Infisical/agent-vault.git
cd agent-vaultUse the official installer, set a master password, and start the daemon.
curl --proto '=https' --proto-redir '=https' --tlsv1.2 -fsSL https://get.agent-vault.dev | sh
export AGENT_VAULT_MASTER_PASSWORD=<your master password>
agent-vault server -dAfter creating a vault, credentials, and an agent token in the UI or CLI, launch your agent via the proxy.
agent-vault run -- claudeIt is built and maintained by Infisical, the open-source secrets management company.
This project is open source and self-hostable at no cost; check the repository license file for exact terms, since the catalog metadata lists it as NOASSERTION. Infisical also sells a commercial, platform-integrated version.
Agent Vault does not need an LLM API key itself; it stores and brokers your existing keys for services like Anthropic or GitHub.
Generic forward proxies such as mitmproxy or Squid move traffic but do not broker credentials; Agent Vault is purpose-built to substitute and guard credentials for agent workflows.
Install it with the get.agent-vault.dev script or Docker, start the server, create a vault and agent token, then run your agent with agent-vault run.
Real-time voice runtime that keeps AI agents talking, working, and present.
Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harnesses without rewrit…
ByteDance's open-source multimodal AI agent stack: GUI agents that see and operate real desktops
Dev agent multiplexer designed for git worktrees and coding agents.
A fully local Manus-style autonomous agent: browses, codes and speaks — no APIs, no subscription
Battle-tested coding rules for AI agents, distilled from classic programming books.
LibreAIカタログに掲載されたAIエージェントプロジェクトの中で最もスターの多いもの — agent-vaultが合わなかった場合に検討したい、現実的な候補リストです。
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
カタログの中でagent-vaultとスター数が近いAIエージェントプロジェクト — 前後の棚もそのまま見て回れます。
Open-source Computer Use as an MCP service: let any agent drive macOS, Linux, and Windows desktops.
A memory agent for your agent fleet: it curates, reconciles, and expires shared memories across tools.
Google's modular toolkit that lets AI coding agents hunt, reproduce, and patch vulnerabilities autonomously.
Self-managing, unbounded context and long-term memory for coding agents: one session, for life.
Pinvou Agent — open-source desktop AI agent for tools, files, knowledge, workflows and real deliverables
NVIDIA's object-oriented, Pythonic approach to building AI agents.