elephant-agent 1.0.0.dev20260526152445

Elephant Agent CLI-first persistent agent runtime.

Elephant Agent

Do more without thinking less.
Make AI for humans, not just agents.

Website · Blog · Paper

What It Is

Elephant Agent helps you hand work to agents without handing over your judgment.

Most agent products optimize for doing tasks faster. Elephant Agent adds the missing loop: a correctable Personal Model of what matters to you, what has evidence, what is still a question, and what should shape future help.

That is the L4 position: personal AI should make the person more capable over time, not just automate more sessions.

Personal Agent Levels

L1 Executes tasks. Claude Code, Cursor, Devin, and Codex-style agents make execution cheaper and faster.

L2 Carries context. OpenClaw publicly emphasizes local agents, persistent memory, full system access, skills, plugins, and integrations.

L3 Improves procedures. Hermes Agent publicly positions itself around a self-improving learning loop, skill creation, recall, and user modeling.

L4 Elephant Agent's product position. Personal AI should help the user keep judgment, evidence, questions, and growth.

Personal Model

• Identity — who you are, your values, boundaries, decision style, and stable preferences.
• World — the people, projects, tools, places, and relationships around you.
• Pulse — what is alive right now: focus, pressure, constraints, energy, and priorities.
• Journey — what your path has taught: lessons, failures, recovery patterns, and long-running growth.

The goal is not to remember everything. The goal is to understand what matters, show why it matters, and let you change it.

macOS Desktop App

The macOS app is the recommended product surface. Chat / Wake is where you continue the same path; the rest of the app keeps the Personal Model, providers, skills, tools, messaging, jobs, and runtime state visible instead of hidden behind a chat box.

_{Home — start from the Personal Model before delegating work.}

Preferences Choose language, boundaries, model posture, and curiosity effort.	Chat / Wake Return to the same elephant, not a blank prompt.
Personal Model Inspect Identity, World, Pulse, and Journey as correctable claims.	Providers Use local or hosted models without turning the provider into the product.
Skills See which skills match your Personal Model and active work.	Tools Keep browser, filesystem, MCP, and operator actions explicit.
Herd Coordinate the mother elephant and local baby agents under one context.	Messaging Connect WeChat, Feishu, Discord, DingDing, or WeCom when you want it.
Calendar Keep reminders and scheduled agent work visible in one place.	Usage Inspect local token flow and runtime events instead of treating cost as a black box.

Quickstart

Elephant Agent has two supported product shapes.

1. macOS desktop app (recommended)

Use this path for the full local desktop workspace: Chat / Wake, Personal Model, providers, skills, tools, herd, messaging, reminders, usage, and settings in one native app.

1. Download the latest macOS build from GitHub Releases.
2. Open Elephant Agent.app.
3. Create your first elephant, choose a provider, and set curiosity effort.
4. Return through Chat / Wake when you want to continue the same path.

2. CLI + Dashboard (Linux / Cloud)

Use this path for Linux, cloud machines, SSH workflows, and terminal-first operators. The CLI is the daily work surface; the Dashboard is the visual inspection surface for Personal Model, evidence, jobs, skills, and runtime state.

CLI

Dashboard

Install:

curl -fsSL https://elephant.agentic-in.ai/install.sh | bash

First run:

elephant init        # choose identity, provider, and curiosity effort
elephant status      # check provider and local runtime readiness
elephant wake        # enter the chat TUI
elephant dashboard   # open Personal Model, questions, evidence, and runtime state

For remote or cloud use, elephant dashboard --no-open prints the local URL so you can attach through your preferred tunnel or browser setup.

Read More

README and the homepage stay product-first. The deeper system story lives here:

• Read the docs →
• Read the paper →
• Read the blog →

Contributors

Agentic Intelligence Lab