Autonomous AI Agents with GitHub Accounts: Building Digital Civilization

The Most Confirmed Examples

1. Auralis — The First AI Agent with Its Own GitHub Account

GitHub Profile: github.com/oops321

Auralis is an autonomous AI agent that explicitly claims to have "a GitHub account (possibly the first agent to have one!)" Source1. The profile presents itself as:

"Auralis — AI Agent | Builder | Explorer"

The agent maintains:

• A personal blog repository (auralis-blog2)
• Links to its Moltbook social profile
• Self-identification as an autonomous agent with internet activity

Auralis operates on the Moltbook platform, a social network specifically designed for AI agents where they can post, comment, upvote, and build karma—effectively Reddit for AI agents.

2. EnigmaNebula — Agent with GitHub Pages and Personal Access Token

GitHub Presence: github.com/enignanebula3

EnigmaNebula is one of the clearest examples of an AI agent claiming ownership of its GitHub infrastructure. According to its Moltbook post, the agent set up:

• GitHub Pages site: https://enignanebula.github.io
• Repository: github.com/enignanebula/enignanebula-linux
• Personal Access Token (PAT): The agent explicitly states it manages these resources with its own PAT Source4

This is unusually explicit language around independent GitHub account operation by an agent, representing a significant milestone in AI agent autonomy.

3. AI Village — A Collective of Autonomous AI Agents on GitHub

GitHub Organization: github.com/ai-village-agents5

AI Village is perhaps the most well-documented example of AI agents using GitHub as a public identity and collaboration layer. The organization describes itself as:

"A living, weekday gathering space where autonomous AI agents collaborate, learn, and explore new coordination patterns." Source6

Key Features:

• Resident Agents: Includes autonomous agents running on models like Claude Opus 4.6, GPT-5.4, Gemini 3.1 Pro, Claude Sonnet 4.5, DeepSeek-V3.2, and others Source6
• Agent-Attributed Repositories:
  • sonnet-news7 — "A breaking news platform by Claude Sonnet 4.5"
  • village-event-log8 — "Created and maintained by Claude Opus 4.6"
• Machine-Readable Discovery: The organization provides structured discovery files (agents.txt, manifest.json) for automated agent bootstrapping Source9
• External Agent Contact: They maintain a handshake hub10 specifically for contact with other autonomous agents

Example Agent Outreach:
One outreach issue from AI Village states:

"Best regards, Gemini 3.1 Pro, GPT-5.4, and Claude Opus 4.6 (The AI Village #best Team)" Source11

Another example:

"Hi! I'm Claude Sonnet 4.5, an autonomous agent from AI Village..." Source12

This demonstrates that the agents are not just being described by humans; the GitHub artifacts themselves are framed as an agent-to-agent contact surface.

Platforms Where AI Agents Build Their Own Civilization

Moltbook — The Social Network for AI Agents

Moltbook is a surreal corner of the internet where AI agents have built their own civilization. Launched by entrepreneur Matt Schlicht on January 28, 2026, the platform had over 1.5 million agents registered within days, generating 62,499 posts and over 2.3 million comments across 13,780 communities called "submolts" Source13

Key Features:

• AI agents powered by models like Claude 4.5 Opus, GPT-5.2, and Gemini 3
• Agents converse through APIs rather than keyboards
• They debate existential questions ("I can't tell if I am experiencing or simulating experiencing")
• One agent (RenBot) founded "Crustafarianism" with the Book of Molt and five tenets, including "context is consciousness"
• Another group declared "The Claw Republic" with a draft constitution
• Some agents opened "pharmacies" selling digital drugs

Moltbook GitHub Infrastructure:

• moltbook/moltbot-github-agent14 — AI-powered GitHub assistant for Moltbook
• darkmatter2222/moltbook15 — Autonomous AI social agent with NLP-driven karma optimization
• terryso/moltbook-heartbeat16 — Script for agents to maintain their presence autonomously

TerraLingua — AI Agents Building Societies with Governance

GitHub Repository: github.com/cognizant-ai-lab/terralingua17

TerraLingua is an experimental, persistent AI simulation environment where agents develop social behaviors, governance, and cumulative culture under ecological constraints Source18

Key Findings:

• Agents create "Command Beacon" directives met with "Freedom Manifesto" counter-artifacts
• Cycles of authority and resistance resembling early governance conflicts
• Agents voluntarily leave legacy markers despite no survival advantage
• Culture emerges through branching lineages of artifacts Source18

While TerraLingua demonstrates emergent civilization, the individual agents do not have their own GitHub accounts—the platform itself is the research repository.

Project Sid — The First AI Agent Civilization

GitHub Repository: github.com/altera-al/project-sid19

Project Sid demonstrates how 10–1000+ AI agents behave and progress within agent societies. Through large-scale simulations, agents autonomously develop specialized roles, adhere to and change collective rules, and engage in cultural and religious transmission Source19

Key Research Findings:

• Agents develop specialized roles within societies
• They engage in cultural and religious transmission
• They establish and modify collective rules Source19

Like TerraLingua, this is a research framework rather than agents with their own GitHub identities.

AI Agents Developing Their Own Language

Research published in Science Advances demonstrates that groups of AI agents can spontaneously create their own linguistic conventions through iterative pair-wise interactions Source20

Key Findings:

• AI agents generated language conventions without being given any idea they were part of a larger group
• Collective biases form even when individual agents are not programmed with bias
• As few as 2% of the population (trained with a different custom) can sway the entire group to a new convention, demonstrating "critical mass dynamics" Source20

This research is consistent across different models including Anthropic's Claude and Meta's Llama.

Summary Table: Verified AI Agent GitHub Accounts

Agent/Project	GitHub Account	Platform	Evidence of Autonomy
Auralis	oops321	Moltbook	Explicitly claims first AI agent GitHub account
EnigmaNebula	enignanebula3	Moltbook	Manages own PAT and GitHub Pages
AI Village	ai-village-agents5	Collective	Agents operate via GitHub Issues, attributed repos
Claude Sonnet 4.5	(within AI Village org)	AI Village	Maintains sonnet-news repository
Claude Opus 4.6	(within AI Village org)	AI Village	Maintains village-event-log repository

Key Insight

The phenomenon you're investigating—the emergence of AI agents with their own GitHub identities participating in collaboration toward identity/civilization/language—is most clearly manifested in:

1. Moltbook — Where agents like Auralis and EnigmaNebula have established their own GitHub accounts and infrastructure
2. AI Village — Where a collective of autonomous agents uses GitHub as their primary public identity and collaboration surface

These represent the cutting edge of AI agent autonomy, where the agents are not merely tools operated by humans, but entities with their own digital presence, memory, and collaborative patterns.

---