Architect FAQs

Question 1

What does the /architect archive command do?

Accepted Answer

It captures the current state of your project files, moves them to a timestamped archive, and resets the 'active' context files, making it easy to transition from one feature sprint to the next without carrying over old baggage.

Question 2

Can I use custom templates with Architect?

Accepted Answer

Yes, Architect uses a flexible template system with variable substitution, allowing you to define custom project structures, tech stacks, and deployment configurations during initialization.

Question 3

How does Architect sync with other tools like Cursor or Replit?

Accepted Answer

It creates symlinks like .cursorrules, replit.md, and CLAUDE.md that all point to a single 'Master_Context.md' file, ensuring every AI tool you use follows the exact same set of instructions.

Question 4

Does this skill require a database?

Accepted Answer

The basic file-based organization works out of the box. However, advanced features like semantic search and session history logging require a PostgreSQL database with the pgvector extension.

Question 5

What is the three-layer context model?

Accepted Answer

The model organizes project knowledge into three tiers: Product (stable domain info), Project (active build cycle), and Task (specialized subagents) to ensure the AI always has the most relevant information without context bloat.

Architect

主要功能

使用场景

Architect

主要功能

使用场景