rebuild 0.1.7

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

rebuild

AI Cost Tracking

• 🤖 LLM usage: $1.0500 (7 commits)
• 👤 Human dev: ~$220 (2.2h @ $100/h, 30min dedup)

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

---

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

---

What it does

1. Walk — iterates through git history day by day (configurable range)
2. Deploy — starts the service per commit (docker-compose / uvicorn / none)
3. Scan — detects endpoints via deta scan → OpenAPI → Traefik labels fallback
4. Test — runs testql scenarios or falls back to HTTP probe per endpoint
5. Screenshot — Playwright screenshots per endpoint with retry/timeout
6. Report — generates report.html + results.json per day + timeline index.html
7. Restore — extracts the last working day of an endpoint as an isolated project
8. Dashboard — overlays health% timeline with avg cyclomatic complexity (via toon)

---

Install

pip install rebuild

# with Playwright screenshots
pip install "rebuild[screenshots]"
playwright install chromium

# development
pip install -e ".[dev]"

---

Quick start

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

# Full walk with docker-compose
rebuild walk . --days 30 --deploy docker-compose

# Walk a specific date range
rebuild walk /path/to/repo --from 2024-01-01 --to 2024-04-01

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

# Generate timeline report from existing results
rebuild report

# Generate CC vs health% dashboard
rebuild dashboard --repo .

# Show version
rebuild version

---

CLI reference

rebuild walk

rebuild walk [REPO] [OPTIONS]

  REPO                    Path to git repository (default: .)

Options:
  --days INT              How many days back (default: 30)
  --from TEXT             Start date YYYY-MM-DD
  --to TEXT               End date YYYY-MM-DD
  --output PATH           Output directory (default: .rebuild)
  --deploy TEXT           Deploy method: auto|docker-compose|uvicorn|none
  --health-url TEXT       Health check URL (default: http://localhost:8003/api/health)
  --base-url TEXT         Base service URL (default: http://localhost:8003)
  --screenshots / --no-screenshots
                          Take Playwright screenshots (default: true)
  --dry-run               Scan only, no deploy

rebuild restore

rebuild restore ENDPOINT [REPO] [OPTIONS]

  ENDPOINT                Endpoint path e.g. /api/health

Options:
  --output PATH           Target project directory (default: restored/)
  --results-dir PATH      Walk results directory (default: .rebuild)

rebuild report

rebuild report [OPTIONS]

Options:
  --results-dir PATH      Walk results directory (default: .rebuild)

rebuild dashboard

rebuild dashboard [OPTIONS]

Options:
  --results-dir PATH      Walk results directory (default: .rebuild)
  --repo PATH             Repository path for CC extraction via toon (optional)

---

Output structure

.rebuild/
  2024-03-15/
    commit.txt              # sha · message · author · timestamp
    endpoints.json          # detected endpoints
    results.json            # HTTP / testql results per endpoint
    testql-results.json     # raw testql output (if testql available)
    report.html             # per-day HTML report with screenshots
    screenshots/
      GET_api_health.png
      GET_api_items.png
  index.html                # timeline of all days
  dashboard.html            # CC vs health% chart (after rebuild dashboard)

---

Configuration — rebuild.yaml

project:
  name: "my-api"
  repo: "."

deploy:
  method: docker-compose          # docker-compose | uvicorn | custom | none
  compose_file: docker-compose.yml
  health_url: http://localhost:8003/api/health
  health_timeout: 60
  health_interval: 2

scan:
  days: 30
  earliest_commit_per_day: true

testql:
  scenarios_dir: testql-scenarios/
  base_url: http://localhost:8003
  timeout: 10

output:
  dir: .rebuild
  screenshots: true
  html_report: true

---

Architecture

rebuild/
├── rebuild/
│   ├── __init__.py
│   ├── cli.py              # Typer CLI: walk, restore, report, version, dashboard
│   ├── models.py           # Dataclasses: WalkConfig, DayResult, Endpoint, CommitInfo
│   ├── git_walker.py       # Day-by-day git history iteration
│   ├── deployer.py         # Deploy method detection + docker-compose/uvicorn lifecycle
│   ├── endpoint_scanner.py # deta scan → OpenAPI → Traefik labels fallback
│   ├── tester.py           # testql runner with HTTP probe fallback
│   ├── screenshotter.py    # Playwright screenshots with retry/timeout
│   ├── reporter.py         # HTML + JSON report per day + timeline index
│   ├── restorer.py         # Extract last working endpoint as isolated project
│   └── dashboard.py        # CC (toon) vs health% comparative timeline
├── tests/                  # 44 tests (pytest)
├── examples/
│   ├── 01-dry-run-walk/
│   ├── 02-docker-compose-project/
│   └── 03-restore-endpoint/
├── testql-scenarios/
├── pyproject.toml
└── rebuild.yaml.example

Endpoint scanner — 3-level fallback

Level	Source	What it returns
1	deta scan	services → ports → health paths
2	/openapi.json	full endpoint list from OpenAPI spec
3	docker-compose Traefik labels	PathPrefix rules
4	hardcoded fallback	GET /api/health

Tester — 2-level fallback

Level	Condition	Behaviour
1	testql installed + testql_dir set	runs .testql.yaml scenarios
2	fallback	HTTP GET probe per endpoint

---

Module summary

Module	CC̄	Functions	Purpose
cli	6.3	8	CLI entry points
models	—	—	Dataclasses only
git_walker	3.3	6	Git history iteration
deployer	3.5	9	Deploy lifecycle
endpoint_scanner	5.5	6	Endpoint detection
tester	4.3	6	Endpoint testing
screenshotter	3.6	5	Screenshots
reporter	6.0	6	HTML/JSON reports
restorer	5.5	5	Endpoint extraction
dashboard	4.5	4	CC vs health chart

Hotspots (CC ≥ 9): walk (CC=15), report (CC=13), extract_endpoint (CC=9), save_html (CC=9), _scan_via_compose_labels (CC=9)

---

Dependencies

Runtime: typer>=0.12, rich>=13, gitpython>=3.1, httpx>=0.27, pyyaml>=6, pydantic>=2, deta>=0.1

Optional: playwright>=1.44 (screenshots)

Dev: pytest>=8, pytest-cov, pytest-asyncio, ruff, mypy

---

Examples

See examples/ for ready-to-run scenarios:

• 01-dry-run-walk/ — scan history without deploy
• 02-docker-compose-project/ — full walk with docker + screenshots + dashboard
• 03-restore-endpoint/ — restore last working endpoint as isolated project

---

License

Licensed under Apache-2.0.