Quick Start Guide
Relevant source files
The following files were used as context for generating this wiki page:
The Quick Start Guide provides a comprehensive roadmap for integrating the Engineering Platform (eng-platform) into both new and existing software repositories. The platform serves as a unified foundation of engineering standards, providing shared configurations (ESLint, Prettier, TypeScript), automated workflows via ÆtherLight skills, and specialized Model Context Protocol (MCP) servers for task management and team analytics.
Sources: README.md:1-15, skills/initialize/SKILL.md:7-15
Core Installation Methods
The platform supports three primary integration strategies depending on the project's maturity and existing tooling.
1. New Project Initialization
For new repositories, the /initialize skill automates the creation of a standardized folder structure, Git workflow, and development standards. This process establishes core directories such as .claude/skills, .aetherlight/patterns, and sprints.
Sources: skills/initialize/SKILL.md:17-48, README.md:28-40
2. Integration into Existing Projects
For established projects, the platform uses a "safe-sync" strategy that prioritizes non-destructive integration. It analyzes existing configurations (ESLint, Prettier, etc.) before suggesting merges, ensuring production code remains stable.
Sources: ADD-TO-EXISTING-PROJECT-PROMPT.md:11-30, PROMPT-ADD-TO-EXISTING.md:4-10
3. LLM-Driven Setup (Claude Code)
Engineers can utilize specific prompts to delegate the setup to AI agents like Claude Code. These prompts automate the addition of eng-platform as a git submodule and the configuration of environment-specific files.
Sources: ADD-TO-REPO-PROMPT.md:5-35, PROMPT-ADD-TO-EXISTING.md:12-25
Architecture and Data Flow
The integration architecture relies on a git submodule established at .eng-platform/. This allows individual projects to stay updated with the central platform standards while maintaining local overrides.
Setup Flow Diagram
The following diagram illustrates the automated initialization sequence triggered by the /initialize command or the LLM prompts.
Components and Configurations
Model Context Protocol (MCP) Servers
MCP servers provide specialized AI capabilities. The platform categorizes these into core, database, and optional servers.
| Server ID | Purpose | Port/Path |
|---|---|---|
beads-integration | Task and subtask management | mcp-servers/beads-integration/ |
team-analytics | Engineer profiles and health metrics | mcp-servers/team-analytics/ |
research-swarm | Multi-step research with confidence tracking | Port 3012 |
basin-analyzer | Context drift detection & AI quality control | Port 9383 |
smart-memory | Cross-platform team knowledge graph | mcp-servers/smart-memory/ |
Sources: PROMPT-ADD-TO-EXISTING.md:29-55, README.md:135-155, mcp-servers/README-SETUP.md:10-35
Standard Configuration Files
The platform enforces uniformity through several base configuration files located in the submodule.
- ESLint: Referenced via
./.eng-platform/configs/eslint/base.js. - Prettier: Copied from
.eng-platform/configs/prettier/.prettierrc.json. - TypeScript: Extends
.eng-platform/configs/typescript/tsconfig.base.json. - Git Hooks: Managed via Husky and lint-staged, using conventional commits (commitlint).
Sources: ADD-TO-REPO-PROMPT.md:10-25, skills/initialize/initialize-eng-platform.md:48-85
Deployment and Team Sync
Once initialized, the platform uses a specific workflow to maintain team knowledge and environment consistency.
Team Memory Synchronization
Team knowledge is persisted in .shared/team-memory.json. This file is separate from the submodule and is committed directly to the project repository to allow team-wide memory sharing.
Environment Variables
Several MCP servers require specific environment keys to function:
VOYAGE_API_KEY: Required forbasin-analyzersemantic embeddings.DEEPSEEK_API_KEY: Required forbasin-analyzerLLM operations.MEMORY_FILE_PATH: Absolute path to the team memory JSON.
Sources: PROMPT-ADD-TO-EXISTING.md:100-105, mcp-servers/README-SETUP.md:125-140
Conclusion
The Quick Start process transforms a standard repository into an AI-native engineering environment. By combining submoduled configurations with localized team memory and specialized MCP servers, the platform ensures that engineering standards are enforced automatically while providing agents with the context needed for autonomous execution. Users are encouraged to start with the /initialize command or the provided LLM prompts to ensure a compliant setup.
Sources: skills/initialize/initialize-eng-platform.md:155-175, README.md:200-215