Axiom Refract vs. Wiki Documentation
The wiki says the architecture works one way. The code says otherwise.
Architecture documentation in wikis, Confluence pages, and Notion databases represents what someone believed the architecture looked like when they wrote it. It does not update itself, cannot verify its own accuracy, and drifts from reality with every merged PR.
Feature Comparison
| Feature | Axiom Refract | Axiom Refract vs. Wiki Documentation |
|---|---|---|
| Architecture Governance | ✓ | — |
| SPOF Detection | ✓ | — |
| Blast Radius Analysis | ✓ | — |
| Dead Code Detection | ✓ | — |
| Dependency Mapping | ✓ | — |
| Compliance Mapping | ✓ | — |
| MCP/AI Agent Integration | ✓ | — |
| Multi-Language (145+) | ✓ | — |
| C4 Diagram Generation | ✓ | — |
| Supply Chain Audit | ✓ | — |
Where This Approach Falls Short
- Wiki documentation is written by humans and never verified against the actual codebase
- No automatic updates — documentation drifts from reality with every code change
- No structural analysis, risk quantification, or compliance mapping capabilities
What Axiom Refract Does Differently
Verified vs. Described
Axiom extracts architecture from code via AST parsing. Wikis describe architecture from memory. One is verified; the other is hopeful.
Always Current
Each Axiom scan reflects the current state of the codebase. Wiki pages reflect the state of the codebase when someone last updated them, which may have been months ago.
Programmatic Access
Axiom data is queryable via API and MCP. Wiki content is human-readable text that cannot be consumed by CI/CD pipelines or AI agents.
Who Should Consider Axiom Refract
Teams that maintain architecture wikis and want to replace (or supplement) manually written documentation with automatically generated, code-verified architectural records.
See it in action.
Upload your repository and get a complete architectural record. No credit card required.