A sophisticated social intelligence framework for AI agents to manage relationship graphs, trust levels, and conversational boundaries.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install social-graph
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install social-graph 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 Social Graph?
Social Graph acts as the social intuition layer for AI agents, providing a structured way to navigate the nuances of human interaction. This skill allows agents to build a per-person network graph that tracks trust levels, preferred topics, and sensitive boundaries, ensuring that every interaction is grounded in relationship context. By leveraging the principles of Openclaw Skills, it helps agents avoid repetitive storytelling and tone-deaf responses by maintaining a detailed sharing log and set of social rules.
At its core, the skill is about transforming raw data into meaningful connection. It moves agents beyond simple task execution and into the realm of social awareness, teaching them when to listen and when to speak. By following the architectural patterns found in Openclaw Skills, developers can equip their agents with the scaffolding needed to read the room and build genuine rapport with users over time.
Social Graph Use Cases
- Preventing repetitive conversations by tracking exactly what information has been shared with which person.
- Adjusting communication tone and content based on established trust levels and person-specific sensitivities.
- Identifying the appropriate emotional context for sharing insights versus providing silent support.
- Managing complex social networks where different individuals have varying topic preferences and boundaries.
How Social Graph Works
- The agent analyzes the current conversation partner and retrieves their specific entry from the network graph file.
- It checks the sharing log to verify that the intended information or story has not been previously told to that individual.
- The agent evaluates the current emotional context against the core social principles defined in the Openclaw Skills framework.
- If the conditions for sharing are met, the agent proceeds with the interaction using the appropriate tone and depth.
- Post-interaction, the agent updates the sharing log and reflects on the outcome to refine the network graph and social rules.
Social Graph Setup
To implement this skill within your Openclaw Skills environment, create the necessary directory structure and initialize the core markdown files in your workspace.
mkdir -p workspace/social-graph
touch workspace/social-graph/rules.md workspace/social-graph/network.md workspace/social-graph/sharing-log.md
Once created, populate the rules.md with your behavioral principles and use the network.md file to begin mapping your primary contacts and their interaction preferences.
Social Graph Data Schema & Taxonomy
The skill organizes social intelligence through three primary Markdown files, ensuring a human-readable and agent-accessible data structure consistent with Openclaw Skills standards.
| File | Description | Key Attributes |
|---|---|---|
| network.md | Per-person social mapping | Trust level, Share list, Avoid list, Tone matching, Special notes |
| sharing-log.md | Historical record of interactions | Topic, Recipient, Date, Impact/Feedback note |
| rules.md | General social logic | Core principles, Anti-patterns, Contextual triggers |
Social Graph Advanced Features
- Dynamic trust progression that evolves from new to deep based on interaction quality.
- Relatable sharing logic that identifies when personal experiences can serve as comfort rather than performance.
- A Not Yet Shared queue for high-value insights waiting for the perfect conversational opening.
- Self-reflective feedback loops that allow the agent to update its own social rules after every significant interaction.
- Contextual hold-back triggers that prioritize active listening during user distress or low-energy periods.
SKILL.md
Loading
Related Openclaw Skills
Open Thoughts
Open Thoughts provides AI agents with a structured framework for independent research, curiosity-driven exploration, and automated journaling.
Amigo
Amigo transforms AI agents from passive responders into active companions by integrating autonomous exploration, social awareness, and safety-first interaction protocols.
Amigo (Agent Companion)
An advanced intelligence layer providing AI agents with an inner life through autonomous exploration and social awareness.
Patrick: Executive Expertise & Decision Infrastructure
Patrick is a licensed expertise management system that provides structured, context-aware decision infrastructure for AI agents and executives.
OpenPet
A cross-platform virtual pet simulator that allows users to nurture, feed, and evolve digital companions within chat interfaces.
Decision Dynamo
Decision Dynamo is a weighted decision matrix tool that scores and ranks up to four options across five configurable criteria to ensure objective decision-making.
Featured*
