rebuild 0.1.10

Historical deployment analysis — walk git history, deploy per day, test all endpoints, capture screenshots, restore working fragments

rebuild

AI Cost Tracking

• 🤖 LLM usage: $1.5000 (10 commits)
• 👤 Human dev: ~$249 (2.5h @ $100/h, 30min dedup)

Generated on 2026-05-01 using openrouter/qwen/qwen3-coder-next

---

Code Evolution Intelligence Engine

Historical deployment analysis & Code Intelligence — walk git history day by day, deploy per commit, test all endpoints, capture screenshots, and analyze code evolution to find duplicates, rank quality, and generate refactor plans.

---

What it does

1. Intelligence Layer (NEW) — Detects structural duplicates, builds service graphs, and ranks code quality across history.
2. Decision Engine (NEW) — Generates actionable refactoring plans (Merge/Extract/Split) with clear rationale.
3. Walk — Iterates through git history day by day (earliest commit per day).
4. Deploy — Starts the service per commit: docker-compose / uvicorn / none.
5. Scan — deta (service-level) → OpenAPI (API-level) → Traefik labels fallback.
6. Test — testql scenarios (detailed API) or HTTP probe fallback (service-level).
7. Screenshot — Playwright screenshots per endpoint with retry/timeout.
8. Restore — Extract last working endpoint as isolated Docker project.
9. Dashboard — Health% timeline overlaid with quality metrics.
10. TUI — Interactive terminal menu for project exploration and recovery.

---

Install

pip install rebuild

# with full intelligence suite (recommended)
pip install "rebuild[full]"
playwright install chromium

---

Quick start

🚀 Analysis & Intelligence (New in v0.1.10)

# Find structural and semantic duplicates
rebuild analyze duplicates .

# Visualize service architecture graph and detect cycles
rebuild analyze services

# Find the 'best' historical version of a function
rebuild analyze truth rebuild/application/pipeline.py run_day

# Generate a prioritized refactoring plan
rebuild refactor plan .

🏃 Execution Pipeline

# Interactive TUI
rebuild tui

# Dry-run walk: scan last 30 days without deploy
rebuild walk . --days 30 --dry-run

# Restore last working version of an endpoint
rebuild restore /api/health .

# Generate health vs quality dashboard
rebuild dashboard --repo .

---

Examples

Explore ready-to-run scenarios in examples/:

• 01-dry-run-walk: Standard walk without deployment + intelligence analysis.
• 02-docker-compose-project: Full pipeline with Docker isolation, screenshots, and architectural refactor plan.
• 03-restore-endpoint: Discovering historical "truth" and extracting a working endpoint as a standalone project.

To run any example:

cd examples/01-dry-run-walk
./run.sh ../.. 30

---

Architecture

Rebuild follows a strict layered architecture: Interfaces → Application (Pipeline/Services) → Intelligence (Analysis/Refactor) → Domain → Infrastructure.

Detailed documentation: docs/architecture.md

rebuild/
├── analysis/           # Intelligence layer: duplication, truth ranking, service graph
├── application/        # Application logic: Pipeline and injected Services
├── domain/             # Pure data models (Endpoint, DayResult, CommitInfo)
├── infrastructure/     # External adapters (Git, HTTP, Playwright)
├── interfaces/         # Entrypoints: CLI, TUI, Dashboard
└── refactor/           # Decision engine: Refactor recommendation engine

---

Changelog & Roadmap

• See CHANGELOG.md for the latest updates.
• See TODO.md for the future roadmap (Automated Refactoring, LLM assistance).

---

License

Licensed under Apache-2.0.