Document shared evidence schema

This commit is contained in:
root 2026-01-02 18:02:10 +00:00
parent cda9700172
commit f05a4e9b1a

40
SCHEMA.md Normal file
View file

@ -0,0 +1,40 @@
# Evidence schema (v1)
This repo is coordinated across parallel sessions. Each session writes its own evidence under:
- `data/ho36/`
- `data/flaneur/`
## `evidence.json`
Top-level:
```json
{
"hostel_name": "string",
"collected_at": "ISO datetime",
"collector_session": "A|B|C",
"evidence": [],
"profile": {}
}
```
Each evidence item (one “fact”) is a row/object with:
- `target`: `"ho36"` | `"flaneur"`
- `source`: `"official_site"` | `"google_maps"` | `"booking"` | `"hostelworld"` | `"tripadvisor"` | `"instagram"` | `"facebook"` | `"tiktok"` | `"press"` | `"other"`
- `metric_name`: string
- `metric_value`: string | number | null
- `url`: string
- `captured_at`: ISO datetime
- `status`: `"ok"` | `"blocked"` | `"unknown"` | `"error"`
- `confidence`: `"high"` | `"med"` | `"low"`
- `notes`: string
- `screenshot_path`: string | null
## `evidence.csv`
One row per evidence item, with the same fields:
`target,source,metric_name,metric_value,url,captured_at,status,confidence,notes,screenshot_path`