GeoJson : build a Polygon based on Point feature

Dear Stackoverflow team, I'm impressed that after a bunch of hours digging the forum I still can't find any question/answer similar to my problem :

I have a GeoJson with a lot of Points features. I collect all Points (green in my example, see figure below) that follow some specification (distance between each of them)

Initial Data:

and I want to link all of them to build a Polygon (which represent an area).

What I'm looking for :

Or Solution accepted :

So I collect all coordinates from these Points, and to be sure the Polygon follows the GeoJson requirements, I'm using the "rewind" function

from geojson_rewind import rewind

But at the end, whatever I've tried I only get that kind of Polygon:

I probably don't use correctly the "rewind" function?

I'm looking for an (easy) automatic way to link all points together in a "convexion hull"

Thanks a lot for any help !

My initial coordinates are collected in a list :

[[4.3556672, 50.8538851], [4.3542534, 50.8546955], [4.3567798, 50.8547854], [4.3566527, 50.8541356], [4.3574286, 50.8552813], [4.3572234, 50.8551264], [4.3547752, 50.8545063], [4.3572736, 50.8560176], [4.3571226, 50.8546104]]

and the Polygon GeoJson I've managed to build, with the rewind function (recopying the last coordinates to get a Polygon) looks like that :

{
  "type": "FeatureCollection",
  "features": [
    {
      "type": "Feature",
      "properties": {},
      "geometry": {
        "type": "Polygon",
        "coordinates": [
          [
            [
              4.357223,
              50.855126
            ],
            [
              4.35678,
              50.854785
            ],
            [
              4.355667,
              50.853885
            ],
            [
              4.356653,
              50.854136
            ],
            [
              4.357123,
              50.85461
            ],
            [
              4.354253,
              50.854695
            ],
            [
              4.354775,
              50.854506
            ],
            [
              4.357429,
              50.855281
            ],
            [
              4.357274,
              50.856018
            ],
            [
              4.357223,
              50.855126
            ]
          ]
        ]
      }
    }
  ]
}

The shapely library is very useful for doing these kinds of geometric manipulations.

For generating a polygon of the convex hull of a set of geometries you can use object.convex_hull:

import shapely.geometry as sg

points = [[4.3556672, 50.8538851], [4.3542534, 50.8546955], [4.3567798, 50.8547854], [4.3566527, 50.8541356], [4.3574286, 50.8552813], [4.3572234, 50.8551264], [4.3547752, 50.8545063], [4.3572736, 50.8560176], [4.3571226, 50.8546104]]
polygon = sg.MultiPoint(points).convex_hull

Which results in the following shape:

Converting it into a GeoJSON with the help of shapely.geometry.mapping:

feature_collection = {
    "type": "FeatureCollection",
    "features": [
        {"type": "Feature", "properties": {}, "geometry": sg.mapping(polygon)}
    ],
}

import json
geojson = json.dumps(feature_collection)