Welcome to GeNCorE Command Center

Relevant source files

The following files were used as context for generating this wiki page:

• README.md
• app/page.tsx
• app/task/page.tsx
• app/review/page.tsx
• lib/agents.ts
• middleware.ts
• app/api/task/route.ts
• app/api/prs/route.ts
• app/api/merge/route.ts

GeNCorE Command Center is a specialized web-based interface designed to manage AI agents, specifically GitHub Copilot, Claude, and Codex. It serves as a central hub for tasking these agents and managing the resulting pull requests within the GentelmeN-CorE repository ecosystem. The system is built with Next.js 15 and leverages the GitHub REST API to orchestrate workflows between human users and autonomous agents.

Sources: README.md:1-5, package.json:2-10

System Architecture and Core Components

The application is structured as a Next.js App Router project with a clear separation between client-side dashboards and server-side API routes that interface with GitHub.

Core Components

• Dashboard (/): Provides a high-level view of agent statuses and active Pull Requests.
• Task Panel (/task): A form-based interface to issue new assignments to AI agents by creating GitHub Issues.
• Review Panel (/review): A dedicated interface for reviewing, approving, and merging pull requests generated by agents.
• Auth Middleware: Protects all routes using a session-based HMAC-SHA256 cookie validation.

Sources: README.md:9-15, app/page.tsx:7-15, middleware.ts:25-50

System Workflow Diagram

The following diagram illustrates the high-level flow from task assignment to pull request management.

AI Agent Ecosystem

The system identifies and manages three specific AI agents. Each agent has unique visual markers and metadata used for task assignment and PR identification.

Agent Profiles

Agent	Identifier	Emoji	GitHub Login	Issue Label	Color Theme
GitHub Copilot	copilot	🟣	copilot	agent:copilot	Purple
Claude (Anthropic)	claude	🟠	claude[bot]	agent:claude	Orange
Codex (OpenAI)	codex	🔵	codex[bot]	agent:codex	Blue

Sources: lib/agents.ts:30-65, README.md:80-85

Agent Detection Logic

The system automatically identifies which agent is responsible for a Pull Request by inspecting the branch name (head.ref) and the author's login.

export function detectAgentFromPR(headRef: string, authorLogin: string): AgentConfig | null {
  const branch = headRef.toLowerCase();
  const author = authorLogin.toLowerCase();

  if (branch.includes('copilot') || author.includes('copilot')) return AGENTS_BY_ID.copilot;
  if (branch.includes('claude') || author.includes('claude')) return AGENTS_BY_ID.claude;
  if (branch.includes('codex') || author.includes('codex')) return AGENTS_BY_ID.codex;
  return null;
}

Sources: lib/agents.ts:85-110

Task Management System

Tasks are initiated through the /task route. When a user submits a task, the system interacts with the GitHub Issues API to create a structured request.

Task Creation Process

1. Selection: User selects an agent (AgentId).
2. Description: User provides a detailed task description.
3. API Execution: The server creates an Issue with the agent as an assignee and applies the specific issueLabel.
4. Local History: The client saves the task to localStorage for quick reference.

Review and Merge Operations

The Review module allows human operators to manage the lifecycle of agent-produced code.

Operation Types

• Approve: Submits a POST request to /api/review with the APPROVE event.
• Request Changes: Prompts the user for a comment and submits a REQUEST_CHANGES event.
• Merge: Executes a PR merge (defaulting to the squash method) via the /api/merge endpoint.

Sources: app/review/page.tsx:45-120, app/api/merge/route.ts:45-60

API Endpoint Specifications

Endpoint	Method	Required Fields	GitHub Action
/api/prs	GET	N/A	Fetches 20 open PRs
/api/review	POST	prNumber, event	Creates a PR review
/api/merge	POST	prNumber	Merges the PR (squash)

Sources: app/api/prs/route.ts:15-30, app/api/review/route.ts:35-50, app/api/merge/route.ts:25-35

Security Implementation

Security is handled via a multi-layered approach to protect both the GitHub credentials and the panel access.

• Authentication: The ADMIN_TOKEN environment variable is required. Access is gated by a middleware.ts that checks for a valid auth cookie.
• Token Protection: The raw ADMIN_TOKEN is never stored in the browser. Instead, an HMAC-SHA256 hash of the token is used as the session identifier.
• Server-Side Secrets: GITHUB_TOKEN is strictly server-side and never exposed to the frontend.
• Input Validation: PR numbers are validated as safe integers to prevent Server-Side Request Forgery (SSRF) when constructing GitHub API URLs.

Sources: README.md:95-105, middleware.ts:30-60, lib/auth.ts:15-30, app/api/merge/route.ts:30-35

Technical Summary

GeNCorE Command Center provides a robust, mobile-optimized (iPhone/Brave) interface for orchestrating AI developer agents. By centralizing tasking via GitHub Issues and PR management via a custom review board, it bridges the gap between high-level human requirements and autonomous agent execution.

Sources: README.md:85-90, app/layout.tsx:10-15