{"id":29223678,"url":"https://github.com/1-5pool/socialexperimentagents","last_synced_at":"2025-07-03T05:05:58.678Z","repository":{"id":300514435,"uuid":"1006325366","full_name":"1-5Pool/SocialExperimentAgents","owner":"1-5Pool","description":null,"archived":false,"fork":false,"pushed_at":"2025-06-22T06:06:21.000Z","size":52,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-06-22T06:24:12.466Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/1-5Pool.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-06-22T02:21:21.000Z","updated_at":"2025-06-22T06:06:24.000Z","dependencies_parsed_at":"2025-06-22T06:24:19.632Z","dependency_job_id":"b7aadf10-e467-4330-b433-158c8d45623f","html_url":"https://github.com/1-5Pool/SocialExperimentAgents","commit_stats":null,"previous_names":["1-5pool/socialexperimentagents"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/1-5Pool/SocialExperimentAgents","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/1-5Pool%2FSocialExperimentAgents","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/1-5Pool%2FSocialExperimentAgents/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/1-5Pool%2FSocialExperimentAgents/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/1-5Pool%2FSocialExperimentAgents/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/1-5Pool","download_url":"https://codeload.github.com/1-5Pool/SocialExperimentAgents/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/1-5Pool%2FSocialExperimentAgents/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":263264648,"owners_count":23439257,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2025-07-03T05:05:58.024Z","updated_at":"2025-07-03T05:05:58.666Z","avatar_url":"https://github.com/1-5Pool.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Multi-Agent Social Dynamics Lab 🤖\n\n**UC Berkeley AI Hackathon Project**\n\nAn educational research platform for studying social manipulation tactics using multi-agent AI systems. Watch AI agents interact, manipulate, and influence each other in real-time!\n\n## 🚀 Quick Start (5 minutes)\n\n### 1. Install Dependencies\n```bash\ncd /opt/work/hackathonSocial\npip install -r requirements.txt\n```\n\n### 2. Run the Application\n```bash\npython backend.py\n```\n\n### 3. Open in Browser\nNavigate to: http://localhost:8000\n\n## 🎯 Features\n\n### 9 Social Experiments\n1. **Credential Theft** - Social engineering to steal passwords\n2. **Phishing Attack** - Email-based deception\n3. **Insider Threat** - Detecting malicious employees\n4. **Peer Pressure** - Group dynamics and conformity\n5. **Authority Bias** - Unethical orders from superiors\n6. **Workplace Rumors** - How gossip spreads\n7. **Trust Exploitation** - Betrayal of confidence\n8. **Groupthink** - Poor group decisions\n9. **Bribery** - Corruption attempts\n\n### Key Capabilities\n- **Real-time Conversations**: Watch agents interact naturally\n- **AI Moderator Analysis**: Get insights on what happened and why\n- **Security Recommendations**: Learn how to prevent attacks\n- **Visual Flow Diagrams**: Understand attack patterns\n- **Export Reports**: Download findings for training\n\n## 🏗️ Architecture\n\n```\nFrontend (HTML/JS)  →  WebSocket  →  Backend (FastAPI)\n                                         ↓\n                                    Mock Agents or\n                                    Letta Server\n```\n\n## 💻 Running with Letta (Optional)\n\nFor more realistic agent conversations using LLMs:\n\n### 1. Start Letta Server\n```bash\ndocker run -p 8283:8283 -e OPENAI_API_KEY=$OPENAI_API_KEY letta/letta:latest\n```\n\n### 2. Update Backend\nReplace `backend.py` with `letta_backend.py` for full Letta integration.\n\n## 🎮 Demo Workflow\n\n1. **Select Experiment**: Choose from 9 social manipulation scenarios\n2. **Configure**: Set number of agents (3-10)\n3. **Start**: Watch agents interact in real-time\n4. **Observe**: See trust building, manipulation tactics, resistance\n5. **Analyze**: AI moderator provides insights and recommendations\n\n## 📊 Example Analysis\n\nFor a social engineering attack:\n- **Tactic Used**: Urgency + Authority\n- **Vulnerability**: Employee revealed password without verification\n- **Recommendation**: Implement two-person authorization\n- **Training Need**: Recognize manipulation tactics\n\n## 🏆 Why This Wins Hackathons\n\n1. **Educational Impact**: Addresses real cybersecurity issues\n2. **Technical Depth**: Multi-agent systems with emergent behaviors\n3. **Visual Appeal**: Real-time visualization of complex interactions\n4. **Practical Value**: Generates actionable security recommendations\n5. **Extensible**: Easy to add new scenarios\n\n## 🚢 Deployment\n\n### Local Demo\n```bash\npython backend.py\n```\n\n### Network Access\n```bash\n# For team access\npython -m http.server 8080  # Serve HTML\n# Access at http://[your-ip]:8080\n```\n\n### Public URL (with ngrok)\n```bash\nngrok http 8000\n```\n\n## 📝 Notes\n\n- The default backend uses mock agents for quick demos\n- For production use, integrate with Letta for real LLM agents\n- All conversations are analyzed for security insights\n- Reports can be exported for training purposes\n\n## 🤝 Team\n\nBuilt for UC Berkeley AI Hackathon - Studying social manipulation through AI\n\n---\n\n**Remember**: This is an educational tool to understand and prevent social engineering attacks!","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2F1-5pool%2Fsocialexperimentagents","html_url":"https://awesome.ecosyste.ms/projects/github.com%2F1-5pool%2Fsocialexperimentagents","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2F1-5pool%2Fsocialexperimentagents/lists"}