diff --git a/gen_resources/README.md b/gen_resources/README.md
new file mode 100644
index 0000000..5ab47a7
--- /dev/null
+++ b/gen_resources/README.md
@@ -0,0 +1,88 @@
+# Generate Dash.jl artifacts
+
+Dash.jl uses Julia
+[Artifacts](https://docs.julialang.org/en/v1/stdlib/Artifacts/) to load
+front-end resources that Dash.jl shares with the python version of
+[dash](https://github.com/plotly/dash).
+
+The [Artifacts.toml](../Artifacts.toml) file lists the location of the
+publicly-available tarball containing all the required resources.
+
+The tarballs are hosted on the
+[DashCoreResources](https://github.com/plotly/DashCoreResources) repo, under
+_Releases_. They are generated and deployed using the `generate.jl` script in
+this directory.
+
+## How to run `generate.jl` ?
+
+### Step 0: get push rights to `DashCoreResources`
+
+### Step 1: get GitHub personal access token
+
+See [GitHub docs](https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/managing-your-personal-access-tokens#creating-a-fine-grained-personal-access-token)
+for more info.
+
+If using a fine-grained token, make sure to enable _Read and Write access to code_.
+
+### Step 2: expose your token into your shell
+
+```sh
+# for example:
+export GITHUB_TOKEN="<your GitHub personal access token>"
+```
+
+### Step 3: run `generate.jl`
+
+```sh
+cd Dash.jl/gen_resources
+
+# install `generate.jl` deps
+julia --project -e 'import Pkg; Pkg.instantiate()'
+
+# generate `gen_resources/build/deploy/` content,
+# but do not deploy!
+julia --project generate.jl
+
+# if everything looks fine,
+# generate `gen_resources/build/deploy/` content (again) and
+# deploy to the `DashCoreResource` releases with:
+julia --project generate.jl --deploy
+```
+
+#### If `generate.jl` errors
+
+<details>
+<summary>with a PyError / PyImport error</summary>
+
+that is an error like:
+
+```sh
+ERROR: LoadError: PyError (PyImport_ImportModule
+
+The Python package dash could not be imported by pyimport. Usually this means
+that you did not install dash in the Python version being used by PyCall.
+
+PyCall is currently configured to use the Python version at:
+
+/usr/bin/python3
+```
+
+try
+
+```jl
+using PyCall
+ENV["PYTHON"] = joinpath(homedir(), ".julia/conda/3/x86_64/bin")
+import Pkg
+Pkg.build("PyCall")
+# check that it matches with
+PyCall.pyprogramname
+```
+
+and then re-run `generate.jl`.
+
+</details>
+
+### Step 4: Commit the changes to `Artifacts.toml`
+
+and push to [plotly/Dash.jl](https://github.com/plotly/Dash.jl)
+(preferably on a new branch) to get a CI test run started.