agen

❯ What's the weather in Taipei?

⏵ Search Tools(weather)

⏵ Plug Tool(Script fetch_weather)

⏺ Currently 28°C in Taipei, partly cloudy with 72% humidity.

Wind from the northeast at 12 km/h.

⏜ 3.2s · claude-sonnet-4-6 · ↑2.1K · ↓340

❯ Report TSMC stock price every morning at 8am on Telegram

⏵ Plug Tool(Script fetch_yahoo_finance)

⏵ Add Schedule(0 8 * * *)

⏺ Done. Scheduled daily at 08:00 — results push to Telegram.

⏜ 5.8s · claude-sonnet-4-6 · ↑3.4K · ↓520

AI Agent Harness

Taiwan-developed AI Agent Harness

Agenvoy is a Taiwan-developed AI Agent Harness by Pardn Chiu, built to turn conversation into completed work on your computer. It coordinates models, context, tools, real-time data, task routing, memory, schedules, and execution in one workflow, while keeping control of your files and environment in your hands. Through the Web interface, it also supports real-time voice conversation you can interrupt mid-sentence, which hands work to the chat agent and narrates its progress aloud; the full result remains available in the chat.

Local-First

Your Data Stays on Your Machine

Meta Muse, OpenAI dots, and Grok Bot each run your agent on a cloud VM the vendor operates, with your logins inside it. Agenvoy runs on the computer you already own, so you decide what leaves it.

Data on Your Disk

Sessions, memory, and schedules live in SQLite under ~/.config/agenvoy. No Agenvoy account, no hosted copy.

Credentials in Your Keychain

API keys and OAuth tokens sit in the OS keychain and are used by the local daemon. The browser page never receives them.

Sandboxed by Default

Commands run under sandbox-exec on macOS or bwrap on Linux with network off unless you approve it for that call. Management endpoints answer localhost only.

Your Choice of Model

Use the provider you trust, or point Agenvoy at Ollama or LM Studio so prompts never leave your machine.

Web Dashboard

Drive Your Agent From the Browser

Manage sessions, tools, schedules, and memory from a browser. The dashboard ships inside the binary — start the daemon and open http://127.0.0.1:17989. It is served by your own machine, so nothing leaves your device.

Features

Everything You Need in One Binary

Auto-built tools and a built-in MCP server at the core — plus multi-model routing, scheduling, and memory, all in a single Go binary.

Auto Tool Generation

Ask for anything. If no tool exists, the agent builds one — writes the script, tests it in a sandbox, and registers it. Next time, it's already there.

MCP Server & Client

Server mode shares every sandboxed tool with Claude Code, Codex, and any MCP client — which can auto-build new tools too. Client mode calls external MCP services.

Multi-Model Dispatch

Eleven providers behind one interface, plus any OpenAI-compatible endpoint. The dispatcher routes each task to the best model automatically.

Scheduler

Cron jobs and one-shot tasks with skill binding. Daemon fires them, hot-reloads on change, auto-disables on failure.

3-Tier Memory

Context window, semantic vector search, full-text SQLite archive. Tool failures and their fixes persist across sessions.

Sandboxed Execution

Every command and script runs inside an OS-native sandbox — bubblewrap on Linux, sandbox-exec on macOS.

Skill System

Loadable markdown instruction packs with YAML frontmatter. Slash-command activation, natural language trigger, or cron-bound.

Self-Improvement

When a tool fails, the agent rewrites it, sandbox-tests it, and retries — the fix and whether it worked carry into later sessions. Every file the tools touch is snapshotted, so a version or a whole task can be restored.

Realtime Voice

The browser dashboard talks with you live through Gemini Live, hands real work to the chat agent, narrates its progress, and lets you cut in mid-sentence. Every request it hands over still lands in the chat transcript.

RAG Integration

Optional KuraDB, registered as an MCP server for external document search. Keyword + semantic dual-track over your personal knowledge base.

How It Works

Capability Gap → Auto-Build

When no tool fits a request, the agent builds one in real-time — finds an API, writes the script, tests it, and saves it for reuse. No code required.

Step 1

Detect Gap

User asks for live data. No matching tool found.

→

Step 2

Find API

Search public APIs, fetch docs, pick the best match.

→

Step 3

Write & Test

Generate script + schema. Run test_tool to verify.

→

Step 4

Call & Share

Execute the new tool. Auto-persisted for all agents.

Claude Code creates a weather tool

via MCP Server

Search local files

File search + RAG

Set up scheduled automation

Cron + Telegram push

MCP

Share Tools Across All Your Agents

One config line gives Claude Code, Codex, and any MCP agent your entire sandboxed tool library — and they can auto-build new tools that everyone instantly shares.

Server

Expose Your Tools

Any MCP-compatible agent connects via stdio and gets access to all tools — built-in, script, API, and extension.

                // claude_desktop_config.json

                {

                  "mcpServers": {

                    "agenvoy": {

                      "command": "agen"

                    }

                  }

                }

Claude Code Codex OpenCode Cursor

Client

Call External Services

Connect third-party MCP servers via stdio or HTTP/SSE. Tools auto-registered and available to the agent immediately.

                // ~/.config/agenvoy/mcp.json

                {

                  "github": {

                    "command": "npx",

                    "args": ["@modelcontextprotocol/server-github"]

                  }

                }

stdio HTTP/SSE Hot Reload

Integrations

Talk to Your Agent Anywhere

Same agent, same tools, same memory — accessible from terminal, chat apps, HTTP, or other AI agents.

TUI

Interactive terminal with slash commands and session switching

Telegram

OTP-authenticated bot with HTML output and voice input

Discord

Guild-verified bot with markdown and button confirms

REST API

OpenAI-compatible chat completions endpoint

MCP Server

Expose tools to Claude Code, Codex, and more

Providers

11 LLM Providers, One Interface

Eleven model providers behind one Agent.Send() interface, plus as many OpenAI-compatible endpoints as you want — Ollama, LM Studio, vLLM or self-hosted. Subscription routes — Claude Code, Codex, Copilot, xAI — are authentication paths, not extra providers. The dispatcher auto-routes each task to the best model.

Anthropic

Opus · Sonnet · Claude Code

OpenAI

GPT-5.x · Codex OAuth

Gemini

Pro · Flash

Copilot

GitHub OAuth

Nvidia

NIM API

xAI Grok

API key · xAI OAuth

DeepSeek

Flash · Pro

Mistral

API key

Ollama Cloud

API key

OpenRouter

200+ models

Cloudflare

Workers AI

+ OpenAI-compatible

Ollama · LM Studio · vLLM

Tools

32 Built-in Tools, Infinitely Extensible

Related actions sit behind one name and a mode argument instead of a dozen near-identical tools. 27 always register; pkg_manage is added on Linux and macOS, and four more once their credential or setting exists. A run carries only find_tools and run_tool; every other tool is listed by name, its schema fetched through find_tools and called through run_tool, so the tool block never changes mid-session and the prompt cache survives. Missing a tool? The agent builds it.

Files

Locate, read, edit and roll back files. Every change is recorded, so one version — or an entire task — can be put back.

find_filesread_files edit_filefile_history

Tools & Skills

Discover what already exists, author new tools and skills, sandbox-test one before it goes live.

find_toolsrun_tooledit_tooltest_tool run_skilledit_skill

Web & Network

Web results and news headlines in one call, a page as markdown, or a raw HTTP request with multipart upload.

search_webfetch_pagehttp_requesthtml_template

Execution

Run a binary under sandbox constraints, hand a file to the OS, fetch an asset, drive the Linux package manager outside the sandbox.

run_commandopen_file download_filepkg_manage

Coordination

Delegate a subtask to its own session, queue timed and recurring runs, keep a live checklist, pause for an answer.

subagentsschedules write_todowrite_resultask_user

State & Memory

The session's own action log, tool failures kept across sessions with what fixed them, reasoning rules pulled in by topic.

chat_historyerror_historyreasoning_guide

Support & Conditional

Arithmetic and unit conversion, keychain-stored secrets. Speech synthesis, image generation and chat push register only once their credential or setting exists.

calculatestore_secretgenerate_audio generate_imagelist_chatbotsend_to_chatbot

Extensibility

Auto-generate tools from natural language. Drop scripts, API descriptors, or connect an MCP server — KuraDB RAG arrives this way.

script_*api_* ext_*mcp__*

Get Started

Up and Running in 60 Seconds

One command to install, one command to start. Add a provider and you're ready.

1

Install

Single binary. macOS and Linux.

curl -fsSL https://agenvoy.com/scripts/install.sh | bash

2

Launch

TUI opens. Daemon auto-starts.

agen

3

Add Provider

In the TUI. Any of 11 providers; the key goes to your keychain.

/model

4

Create Session

Name it. Persists across restarts.

/new

New here? Free key to first answer in 5 minutes →

Your 24/7 Personal AI AgentRunning on Your Own Machine