seo-log-auditor

Streamlit dashboard that ingests a 30-day Grafana/Loki export of nginx-ingress Googlebot logs and runs seven SEO crawl-audit techniques on it:

1. Crawl budget distribution – how hits are spread across page types vs how URLs are spread across page types.
2. Orphan pages – URLs Googlebot is hitting that are not in your sitemap.
3. Status-code waste – non-200 ratio per page type, redirect chains.
4. Stale high-value pages – sitemap URLs not crawled in the last N days.
5. Performance – page-size vs latency vs hit frequency.
6. Bot verification – verified Googlebot vs spoofed User-Agents.
7. Parameter traps – paths with an explosive number of query-string variants.

Quick start

Zero-install via uv:

uvx seo-log-auditor

Or install into the current environment:

pip install seo-log-auditor
seo-log-auditor

Either command launches the dashboard at http://localhost:8501. To pass flags through to Streamlit:

seo-log-auditor --server.port 9000 --server.headless true

Then in the sidebar:

1. Upload your Grafana/Loki export (.json, .csv, or .txt).
2. Paste your sitemap URL (a sitemap index works too).
3. Optionally upload a page_patterns.yaml (see src/seo_log_auditor/config/page_patterns.example.yaml).

Exporting logs from Grafana

Use this LogQL query in Grafana Explore against your Loki datasource:

{app="ingress-nginx"} |= "Googlebot"

Set the time range to last 30 days and download as JSON (best fidelity) or CSV. Plain .txt (one log line per row) also works.

Page-pattern config

src/seo_log_auditor/config/page_patterns.example.yaml is the starting point. First match wins. Edit it for your URL structure and re-upload via the sidebar. Tagging traps like paginated and faceted as their own page types makes leakage immediately visible on the Crawl Budget page.

Development

git clone https://github.com/hitensangani/seo-log-auditor.git
cd seo-log-auditor
uv sync --extra dev
uv run streamlit run src/seo_log_auditor/app.py

Project layout

src/seo_log_auditor/
  app.py                        # Streamlit entry, sidebar uploads, KPI overview
  cli.py                        # `seo-log-auditor` console script
  pages/                        # Multipage app, one file per technique
  config/                       # Example page-pattern rules
  parsers.py                    # Loki JSON / CSV / nginx-text -> DataFrame
  classify.py                   # Regex-based URL -> page_type
  sitemap.py                    # Fetch + parse sitemap.xml (incl. index)
  verify_bot.py                 # Google IP ranges + cached reverse DNS
  analysis/                     # One module per technique
tests/                          # pytest fixtures + unit tests

Running tests

uv run pytest

Roadmap

• Internal-link-depth correlation (technique 4 deeper layer) once you have a Screaming Frog / Sitebulb export.
• Direct Loki API streaming so you don't need to download files.
• SQLite cache for week-over-week comparisons.

License

MIT — see LICENSE.