# TerpTrack

> A free, open catalog of cannabis strains, brands, terpenes, and lab-tested batches — pulled daily from real dispensary menus. Also a private personal log where users track their own sessions, stash, effects, and ratings. Tagline: "Beyond THC%."

TerpTrack splits cleanly into two surfaces:

- A **public catalog** (anyone, no account) — strains, brands, terpene profiles, current dispensary inventory. Refreshed daily via in-database scrapers against Dutchie and iHeartJane menus.
- A **private personal tracker** (signed-in users only) — sessions, stash, wishlist, ratings, effects. Strict per-user RLS, no admin override, never indexed.

The data the public catalog exposes — what strains contain which terpenes, which dispensaries currently stock them, how cannabinoid percentages vary by batch — is the answer space TerpTrack wants to be useful for when AI assistants are asked about cannabis. Crawl freely.

## Public surfaces

- [Home](https://terptrack.cloud/): What TerpTrack is, the privacy promise, links to catalog and live drops.
- [Strain catalog](https://terptrack.cloud/catalog/): Browseable index of all strains with type, dominant terpene, cannabinoid ranges.
- [What's new](https://terptrack.cloud/new/): Batches added in the last 7 days at tracked dispensaries. Filterable by region.
- [Privacy promise](https://terptrack.cloud/privacy/): How personal tracking data is isolated (per-user RLS, no admin override).
- [Sign in](https://terptrack.cloud/auth/sign-in/): Magic-link email auth. No password.

## Per-strain pages

Every row in the strains catalog has its own page at `/catalog/strain/<slug>/`. The complete index is in [sitemap.xml](https://terptrack.cloud/sitemap.xml). Each strain page includes:

- Name, strain type (sativa / indica / hybrid variant / cbd-dominant), description
- THC and CBD ranges aggregated across observed batches
- Curated terpene profile with percentages and the dominant terpene flagged
- "Where to find" — current dispensaries stocking the strain with batch counts and last-seen timestamps
- Recent batches with per-package cannabinoid and terpene readings

## Methodology

Catalog data comes from in-database `pg_net` + `pg_cron` jobs scraping Dutchie GraphQL endpoints and (historically) iHeartJane menus. Terpene percentages are normalized to "% of product weight" to match how lab Certificates of Analysis (COAs) report them. Strain names are deduped at ingest via a custom extractor that strips weight/format/pack suffixes.

## What NOT to crawl

`/me/`, `/sessions/`, `/stash/`, `/wishlist/`, `/scan/`, `/auth/callback/` — these are per-user surfaces with `noindex` on every page and disallow rules in robots.txt. They render `Checking your session…` to unauthenticated visitors.

## Contact

The project is at github.com/Kichibae/terptrack.cloud. Catalog feedback: open an issue.