Skip to main content

System Foundation

Role: Solution Architect | Technical Lead 🏗️

Foundation is the set of foundational technical documents of a System — together they form the Knowledge Base that the AI Agent reads before generating BRD, FRD, or Implementation Plan. The more complete and accurate the Foundation documents, the more the AI output reflects the real tech stack and architecture of the project.

How to use​

  1. In System Management, select the System from the left list.
  2. Switch to the "Foundation" tab.

Foundation tab layout​

The tab is split into two areas:

  • Left column — list of existing Foundation Documents, with a search field and the "+ New Document" button at the top.
  • Right pane — editor for the content of the selected document.

Empty state when there are no documents:

  • Left column: "No documents yet" — "Create your first foundation document"
  • Right pane: "Select a document — Choose a document from the list to start editing"

Create a new Foundation Document​

  1. Click "+ New Document" at the top of the left column.

  2. The "New Foundation Document" modal opens. Fill in the information:

    FieldRequiredNotes
    Document NameYesDocument title (up to 200 characters)
    DescriptionNoShort description of the document's purpose (up to 500 characters, with character counter)

  3. Click "Create". The new document appears in the list and is auto-selected.

Editing content​

After creation, the document is selected and the right editor is shown. The initial content is empty.

  1. Click the Edit icon (pencil) to enter edit mode.
  2. Write the content in the rich text editor (Tiptap) — supports Markdown, headings, bullet lists, code blocks, tables, and more.
  3. When finished, click "Save" to save. Click "Cancel" to discard changes and return to the previous content.

Manage Foundation Documents​

Search: type into the "Search documents..." field to filter by name.

Switch documents: click an item in the left list — its content loads into the right editor.

Each item in the list shows:

  • Document name (bold)
  • Short excerpt (first 120 characters of content)
  • Last updated timestamp

Delete a document: click the trash icon 🗑️ that appears on hover → confirm in the "Delete Document?" modal:

"This will permanently delete "[document name]" and all its content."

Click "Delete" (red) to confirm. This action cannot be undone.

Below is a list of documents commonly created for a technical System. None are strictly required, but the more complete the set, the better the AI Agent performs:

DocumentSuggested contents
Tech StackLanguages, frameworks, and main libraries in use
Architecture OverviewHigh-level diagram, architecture style (microservice, monolith, etc.), main components
Coding StandardsNaming conventions, folder structure, code style guide
Database DesignERD, naming convention, indexing
API Design PrinciplesREST/GraphQL standards, versioning, error format
Security StandardsAuth method (JWT/OAuth), RBAC, input validation
Deployment & DevOpsCI/CD pipeline, Docker, cloud provider, deployment procedure
Performance StandardsSLA, caching strategy, rate limiting
Integration PointsExternal APIs, third-party services, integrated message queues
Prioritize when time is limited

If you do not have time to write all of them, prioritize in this order: Tech Stack → Architecture Overview → API Design Principles → Database Design. These four give the AI enough context to produce good-quality technical documents from the start.

End of Phase 3

Completing this step means the System has been fully initialized. The team can now start Phase 4: AI-Agent-driven artifact generation — producing BRD, FRD, and Test Cases from a Jira ticket, with the AI Agent leveraging all the context you just set up.