Secret & Env Shield
Deterministic regex scanners audit your codebase for accidental private keys, database credentials, and unvalidated environment variables — comparing staging vs production definitions without ever reading secret values.
Deterministic AST parser turning any GitHub repository into an automated architectural audit, interactive dependency graph, and production health score — cutting code review and onboarding time from days to seconds.
Evaluated across 7 architectural dimensions · 82 components analyzed · 0 circular dependencies
Zero unoptimized HTML image tags detected across UI components directory.
Consider splitting view controller logic into subcomponents to maintain testability.
Clean unidirectional module boundary graph between routes, services, and utilities.
Reads the code you already ship
Arbor reads every file through a standard Babel AST parser and constructs an interactive Directed Acyclic Graph of your entire import architecture. Our dependency graph visualization flags forbidden cross-layer imports and cyclic dependencies before they ever reach main.
Every repository receives a 0–100 audit across 7 dimensions — architecture, technical debt, security, documentation and more. The score updates on every run, so a shields badge in your README always reflects the latest audit.
Replaces ambiguous summaries with concrete syntax structures, dependency maps, and Conventional Commits impact.
Deterministic regex scanners audit your codebase for accidental private keys, database credentials, and unvalidated environment variables — comparing staging vs production definitions without ever reading secret values.
Import Swagger or OpenAPI schemas and execute test HTTP requests through a secure Node.js CORS bypass proxy — no external tools, no context switching.
AI-generated codebase documentation derived from the AST itself. Versioned snapshots stay immutable while automatic 90-day staleness alerts keep the docs honest.
Deterministic codebase intelligence without the maintenance tax of manual documentation or the hallucinations of black-box AI.
| Capability | Arbor AST | Manual Wikis (Notion/Confluence) | LLM AI Code Reviewers | CLI Linters |
|---|---|---|---|---|
| Analysis Speed | < 30 seconds | Days / Weeks | 2-5 minutes (slow tokens) | Instant (surface only) |
| Accuracy & Determinism | 100% AST deterministic | Prone to human drift | Hallucination risk | Rule-based syntax only |
| Interactive Dependency Graph | Included | |||
| Zero Source Code Stored | Included | Stored in doc tools | Sent to LLM clouds | Included |
| Dynamic SVG README Badges | Included | |||
| Commit Velocity & Conventional Recency | Included | |||
| Setup Effort | Zero (1-click GitHub OAuth) | Heavy manual authoring | API key & prompts required | Config files (.eslintrc, etc.) |
Arbor is free forever for individual developers and open-source projects. Pro subscriptions launch soon for expanding teams and automated pipelines.
Ideal for solo builders, open source maintainers, and indie hackers.
For growing engineering teams requiring automated audits, team workspaces, and CI/CD gates.
Clear answers about security, GitHub permissions, and the AST engine.
Connect a GitHub repository with read-only permissions and get a deterministic architectural audit, dependency graph, and health score — no credit card, no AI hallucination, no source code stored.