Wiki synthesis

Wiki synthesis transforms source code into a structured Functional-Domain Wiki by analyzing your repository's dependency graph, file structure, and code semantics.

Synthesis modes

Domain-partitioned pipeline — the primary mode for production repositories. A 6-stage pipeline that partitions the codebase using community detection, discovers and reconciles domains across partitions, then generates the full hierarchy with 4-5 levels of decomposition.

Hierarchical mode — a single-pass mode for smaller repositories (under ~50 files). Faster but produces a shallower hierarchy.

Stractal selects the appropriate mode automatically based on repository size.

Configuring synthesis

Before starting synthesis, you can adjust settings from Project settings > Inclusion:

Branch — choose which branch to analyze. Defaults to the repository's default branch.
Inclusion patterns — use glob patterns to control which files are analyzed. Useful for excluding test fixtures, generated files, or vendor directories.
Per-repository overrides — for multi-repo projects, set branch and inclusion patterns independently for each repository.

Re-synthesizing

Click Synthesize again after significant code changes. Previous versions are preserved in version history.

Quality gates

Gate	What it checks
Merge rate	Fewer than 15% of domains merged during reconciliation
Flow atomicity	Each leaf business domain is a single atomic operation
Domain count	Proportional to codebase size
Average depth	Hierarchy achieves 4-5 levels

Exploring the Wiki

The Wiki viewer offers two browse modes:

Tree view — full nested hierarchy. Expand business domains to see sub-domains and technical domains.
Category view — domains grouped by strategic category (Core, Infrastructure, Support, External).

Use the search bar to find domains by name or keyword across the entire hierarchy.

Domain details

Select any domain to see:

Artifact	Description
Description	What the domain does and its responsibilities
Business logic	Business rules and invariants (business domains)
Implementation notes	How the domain is built (technical domains)
Owned tables	Database tables this domain is the primary writer for
API endpoints	REST endpoints with method, path, and auth info
Page routes	Next.js routes with URL patterns
Key types	TypeScript interfaces, types, enums, and Zod schemas
State management	React contexts, hooks, and stores
Architecture diagrams	Dependency graphs showing relationships between domains
Dependencies	Other domains this domain depends on or is depended upon by

Exporting the Wiki

You can export the full Wiki from the viewer:

Download as markdown — a .md file containing the complete domain hierarchy with all artifacts
Copy to clipboard — paste the Wiki content into other tools or documents

DDD insights

Wikis synthesized with the domain-partitioned pipeline include Domain-Driven Design artifacts:

Ubiquitous Language — a glossary of terms extracted from your codebase with definitions, source evidence, and DDD classification (entity, value object, aggregate root, domain service, domain event, specification, repository, factory, or module)
Bounded Contexts — explicit model boundaries grouping related domains, with descriptions of model dialect, core abstractions, and relevant vocabulary
Context Map — relationships between contexts showing integration patterns: Shared Kernel, Customer/Supplier, Anticorruption Layer, Conformist, Separate Ways, Open Host Service, Published Language, or Partnership

Access these from the dedicated tabs in the Wiki viewer.

Version history

Each synthesis creates a versioned snapshot. From the version history panel you can:

Browse previous versions — see when each Wiki was synthesized and summary metadata
Compare versions — view a diff showing added, removed, and modified domains between any two snapshots
Restore context — understand how your architecture has evolved over time