Updated Architecture

[nullchimp]

This pull request introduces several significant updates to the project, including the addition of a new prompt for creating architectural decision records (ADRs), enhancements to existing prompts to integrate ADR references, a rename of the "implementer" persona to "programmer," and a substantial rewrite of the README.md to improve clarity and structure. Below are the most important changes grouped by theme:

Prompt Enhancements

• Architect Persona: Introduced a new architect.prompt.md file to guide the creation of Architecture Decision Records (ADRs). This includes detailed workflows, operating principles, and guardrails to ensure high-quality documentation.
• Planner Persona Updates:
  • Added support for referencing ADRs (${input:adrReference}) in planning workflows.
  • Introduced an architectural review step to assess the need for ADRs during planning and guide users to consult the architect persona when necessary.
  • Updated question-generation and task-creation workflows to incorporate ADR references for better alignment with architectural decisions. [1] [2]

Documentation Improvements

• README.md Overhaul: Reorganized and expanded the README.md to provide a clearer project overview, updated architecture details, and streamlined setup instructions. Added badges for license, Python version, and code style.

Terminology Update

• Renamed implementer.prompt.md to programmer.prompt.md: Updated the persona description to use "programmer" instead of "implementer" for consistency and clarity.

[Copilot]

Pull Request Overview

This PR restructures import paths across tests and core modules, removes outdated documentation, introduces a new Architect persona prompt, enhances the Planner prompt to integrate ADR references, renames the Implementer persona to Programmer, and significantly rewrites the README for clarity and structure.

• Standardized import paths by removing the src. prefix
• Deleted legacy docs and overhauled architecture documentation
• Added ADR support: new architect.prompt.md, planner.prompt.md updates, README badges and sections

Reviewed Changes

Copilot reviewed 27 out of 27 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
tests/*	Updated imports to use top-level libs and core packages
src/core/rag/embedder/*	Adjusted schema imports to new core.rag.schema path
docs/remote-memgraph-connection.md	Removed outdated connection guide
docs/rag_architecture.md	Removed legacy RAG architecture doc
docs/memgraph-troubleshooting.md	Removed old troubleshooting guide
docs/azure-deployment-guide.md	Removed obsolete Azure deployment guide
docs/architecture.md	Completely rewrote with front-matter and updated diagrams
README.md	Simplified structure, added badges, reorganized setup & usage
.github/prompts/programmer.prompt.md	Renamed persona from “implementer” to “programmer”
.github/prompts/planner.prompt.md	Added ADR reference input and architectural review step
.github/prompts/architect.prompt.md	Introduced new ADR-focused Architect persona prompt

Comments suppressed due to low confidence (2)

tests/test_libs_search_fixed.py:4

• The import paths were changed from src.libs... to libs..., but unless the project’s root package is configured accordingly (e.g., via PYTHONPATH or an installation), this will raise ModuleNotFoundError. Please update your test runner configuration or revert to the correct package namespace.

from libs.search.client import GoogleClient, SearchResult, SearchResults

tests/test_core_rag_schema.py:4

• Similar to other tests, this import assumes a top-level core package. Ensure the module path matches your package structure or adjust your project’s import settings.

from core.rag.schema import Document, DocumentChunk, ProcessingStatus, EdgeType, Node, Source, Interaction, VectorStore, Vector

[Copilot]

Typo in 'Stricktly'—should be 'Strictly'.

Suggested change

	* Stricktly consider ${criteria}, ${context}, and any identified ADRs when writing questions.
	* Strictly consider ${criteria}, ${context}, and any identified ADRs when writing questions.