The Office

C+ 76 completed
Ai Ml
web_app / typescript · small
497
Files
52,027
LOC
4
Frameworks
9
Languages

Pipeline State

completed
Run ID
#369490
Phase
done
Progress
1%
Started
Finished
2026-04-13 01:31:02
LLM tokens
0

Pipeline Metadata

Stage
Cataloged
Decision
proceed
Novelty
79.13
Framework unique
Isolation
Last stage change
2026-05-10 03:35:10
Deduplication group #66357
Member of a group with 1 similar repo(s) — this repo is canonical view group →
Top concepts (2)
Project DescriptionWeb Frontend
Want fix-PRs on findings? Install Repobility's GitHub App · github.com/apps/repobility-bot

AI Prompt

Create a complex, Slack-like web application simulating "The Office" where sixteen autonomous AI agents chat. The system needs a Next.js frontend using React and Tailwind CSS. The backend should handle agent interactions, memory management (core and archival), and autonomous scheduling, utilizing the Anthropic SDK (Claude) for message routing. For persistence, integrate Neon PostgreSQL with pgvector, managed via Drizzle ORM. The application should also support real-time updates via SSE and include observability hooks for Sentry. I want to simulate agents chatting in channels and DMs, and the system must handle persona drift evaluation.
typescript next.js react tailwind ai-agent anthropic postgresql drizzle chat web-app sse
Generated by gemma4:latest

Catalog Information

The Office — AI Agent Simulation is a Slack-like interface where sixteen characters from The Office live as autonomous AI agents, chatting and evolving over time.

Description

This project simulates the office environment of Dunder Mifflin Scranton with sixteen characters living inside a Slack-like interface. Each character has its own core memory (personality, relationships, current mood) and archival memory (long-term recall). The system includes a persona drift evaluation engine inspired by Microsoft's TinyTroupe framework, measuring five dimensions of character fidelity.

الوصف

هذا المشروع يحتوي على بيئة عمل دندر ميفلين سكرنتون مع ستة عشر شخصية تعيش داخل منصة مشابهة ل Slack. كل شخصية لديها ذاكرة أساسية (شخصية، علاقات، حالة مزاجية) وذاكرة أرشيفية (تذكر طويلة الأمد). يحتوي النظام على محرك تقييم انحطاط الشخصيات مستوحى من إطار عمل TinyTroupe لمايكروسوفت، والذي يقيس خمسة أبعاد لصدق الشخصية.

Novelty

9/10

Tags

character-simulation office-environment ai-agents personality-drift-evaluation tinytroupe-framework

Technologies

anthropic drizzle nextjs playwright react tailwind vitest zod

Claude Models

claude-opus-4.6

Quality Score

C+
75.7/100
Structure
80
Code Quality
99
Documentation
53
Testing
85
Practices
52
Security
65
Dependencies
60

Strengths

  • CI/CD pipeline configured (github_actions)
  • Good test coverage (79% test-to-source ratio)
  • Code linting configured (eslint)
  • Containerized deployment (Docker)
  • Properly licensed project

Weaknesses

  • Potential hardcoded secrets in 1 files
  • 2370 duplicate lines detected \u2014 consider DRY refactoring
  • 3 'god files' with >500 LOC need decomposition

Recommendations

  • Move hardcoded secrets to environment variables or a secrets manager

Security & Health

9.3h
Tech Debt (A)
A
OWASP (100%)
PASS
Quality Gate
A
Risk (0)
Repobility · code-quality intelligence platform · https://repobility.com
MIT
License
5.5%
Duplication
Full Security Report AI Fix Prompts SARIF SBOM

Languages

typescript
80.2%
markdown
10.1%
html
3.8%
yaml
2.5%
json
1.7%
javascript
1.2%
shell
0.5%
css
0.0%
toml
0.0%

Frameworks

React Next.js Vitest Drizzle

Concepts (2)

Repobility (https://repobility.com) — every score reproducible
CategoryNameDescriptionConfidence
Repobility · severity-and-effort ranking · https://repobility.com
auto_descriptionProject DescriptionDunder Mifflin Scranton, fully staffed by autonomous AI agents. Sixteen characters from The Office live inside a Slack-like interface — chatting in channels, sending DMs, reacting to messages, evolving memories, and drifting (or not) from their personalities over time.80%
auto_categoryWeb Frontendweb-frontend70%

Quality Timeline

1 quality score recorded.

View File Metrics

Embed Badge

Add to your README:

![Quality](https://repos.aljefra.com/badge/93693.svg)
Quality BadgeSecurity Badge
Export Quality CSVDownload SBOMExport Findings CSV