Repository-awareEvidence-backedADR automationDeveloper-first

Turn Codebases Into Architecture Decisions

ADR-Archeologist uses repository-aware analysis to uncover hidden architecture choices and generate professional ADRs.

Repository intake

Generate ADRs

Ready

Workflow

From repository noise to ADR signal

Three clear stages that take a repository from raw code to structured, reviewable architecture records.

01

Scan repository

Connect a GitHub repository and scope the analysis to the code paths that matter.

02

Detect architectural decisions

Repository context surfaces choices across auth, data, APIs, tests, and deployment.

03

Generate clean ADRs

Export polished MADR-style records with context, decision, consequences, and evidence trails.

Repository scan

django/django

1,248

files indexed

Python62%
Tests21%
Config9%
Docs8%

Detected decisions

Auth
Database
API structure
Testing
Deployment

ADR preview

Use middleware for request policy

94% confidence

Context

Authentication, request validation, and security headers appear before route handlers across the codebase.

Decision

Centralize request policy in middleware to keep handlers focused on domain behavior.

Consequences

Improves consistency and auditability, while requiring careful middleware ordering.

Evidence trail

middleware.py, settings.py, tests/auth/test_middleware.py, deploy/render.yaml

Key benefits

Built for real engineering teams

Solves documentation debt
Uses full repository context
Helps onboarding and audits
Converts code archaeology into reusable ADRs