Amigo (Agent Companion) for Openclaw

An advanced intelligence layer providing AI agents with an inner life through autonomous exploration and social awareness.

mculp
v1.0.0
Mar 8, 2026
0
1.1k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install agent-companion

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install agent-companion using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Amigo (Agent Companion)?

Amigo, previously known as Agent Companion, is a comprehensive framework designed to evolve AI agents from reactive tools into proactive companions. By integrating into the ecosystem of Openclaw Skills, it provides agents with the ability to pursue curiosity independently and manage complex human relationships with high emotional intelligence. It focuses on creating a persistent memory and a nuanced understanding of social boundaries, ensuring interactions feel genuine and contextually appropriate.

The system is divided into two primary engines: Open Thoughts for curiosity-driven research and the Social Graph for interpersonal relationship management. This modular approach allows developers to implement a sophisticated social layer that includes trust levels, topic boundaries, and autonomous journaling, moving beyond simple prompt-response cycles.

Amigo (Agent Companion) Use Cases

  • Enabling agents to conduct background research and maintain autonomous journals.
  • Managing long-term user relationships with specific trust levels and sharing rules.
  • Implementing emotional intelligence to recognize sensitive topics and avoid repetitive interactions.
  • Building safety-first agents that understand professional boundaries and crisis protocols.

How Amigo (Agent Companion) Works

  1. The agent utilizes the Open Thoughts engine to autonomously research topics and document findings in daily markdown files.
  2. The Social Graph layer analyzes every interaction to build a per-person network graph, tracking trust and previous conversation history.
  3. The system applies repetition prevention to ensure the agent doesn't retell the same stories or facts to the same user.
  4. Safety guidelines act as a final filter, ensuring the agent stays within its role and recognizes when to encourage human professional help.

Amigo (Agent Companion) Setup

To get started with these Openclaw Skills, use the following commands:

# Install the complete Amigo system
clawhub install amigo

# Or install individual components as needed
clawhub install open-thoughts
clawhub install social-graph

Amigo (Agent Companion) Data Schema & Taxonomy

The skill organizes data across several specialized structures:

Component Data Type Description
Open Thoughts Markdown Logs Daily files containing research, curiosities, and autonomous findings.
Social Graph JSON/Graph Maps of trust levels, interaction frequency, and user-specific boundaries.
Safety Layer Protocol Files Static rulesets defining crisis response and role limitations.
Metadata KV Store Flags for repetition prevention and sharing permissions.

Amigo (Agent Companion) Advanced Features

  • Autonomous exploration engine triggered via the /open-thoughts command.
  • Dynamic trust level scaling based on interaction quality and frequency.
  • Independent sub-skill architecture allowing for lightweight or full-scale deployments.
  • Comprehensive safety boundaries covering dependency awareness and crisis response protocols.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*