Skip to main content

Datasette plugin and ASGI middleware that authenticates users against GitHub

Project description

datasette-auth-github

PyPI Changelog Tests License

Datasette plugin that authenticates users against GitHub.

Setup instructions

  • Install the plugin: datasette install datasette-auth-github
  • Create a GitHub OAuth app: https://github.com/settings/applications/new
  • Set the Authorization callback URL to http://127.0.0.1:8001/-/github-auth-callback
  • Create a metadata.json file with the following structure:
{
    "title": "datasette-auth-github demo",
    "plugins": {
        "datasette-auth-github": {
            "client_id": {"$env": "GITHUB_CLIENT_ID"},
            "client_secret": {"$env": "GITHUB_CLIENT_SECRET"}
        }
    }
}

Now you can start Datasette like this, passing in the secrets as environment variables:

$ GITHUB_CLIENT_ID=XXX GITHUB_CLIENT_SECRET=YYY datasette \
    fixtures.db -m metadata.json

Note that hard-coding secrets in metadata.json is a bad idea as they will be visible to anyone who can navigate to /-/metadata. Instead, we use Datasette's mechanism for adding secret plugin configuration options.

By default anonymous users will still be able to interact with Datasette. If you wish all users to have to sign in with a GitHub account first, add this to your metadata.json:

{
    "allow": {
        "id": "*"
    },
    "plugins": {
        "datasette-auth-github": {
            "...": "..."
        }
    }
}

The authenticated actor

Visit /-/actor when signed in to see the shape of the authenticated actor. It should look something like this:

{
    "actor": {
        "display": "simonw",
        "gh_id": "9599",
        "gh_name": "Simon Willison",
        "gh_login": "simonw",
        "gh_email": "...",
        "gh_orgs": [
            "dogsheep",
            "datasette-project"
        ],
        "gh_teams": [
            "dogsheep/test"
        ]
    }
}

The gh_orgs and gh_teams properties will only be present if you used load_teams or load_orgs, documented below.

Restricting access to specific users

You can use Datasette's permissions mechanism to specify which user or users are allowed to access your instance. Here's how to restrict access to just GitHub user simonw:

{
    "allow": {
        "gh_login": "simonw"
    },
    "plugins": {
        "datasette-auth-github": {
            "...": "..."
        }
    }
}

This "allow" block can be positioned at the database, table or query level instead: see Configuring permissions in metadata.json for details.

Note that GitHub allows users to change their username, and it is possible for other people to claim old usernames. If you are concerned that your users may change their usernames you can key the allow blocks against GitHub user IDs instead, which do not change:

{
    "allow": {
        "gh_id": "9599"
    }
}

Restricting access to specific GitHub organizations or teams

You can also restrict access to users who are members of a specific GitHub organization.

You'll need to configure the plugin to check if the user is a member of that organization when they first sign in. You can do that using the "load_orgs" plugin configuration option.

Then you can use "allow": {"gh_orgs": [...]} to specify which organizations are allowed access.

{
    "plugins": {
        "datasette-auth-github": {
            "...": "...",
            "load_orgs": ["your-organization"]
        }
    },
    "allow": {
        "gh_orgs": "your-organization"
    }
}

If your organization is arranged into teams you can restrict access to a specific team like this:

{
    "plugins": {
        "datasette-auth-github": {
            "...": "...",
            "load_teams": [
                "your-organization/staff",
                "your-organization/engineering",
            ]
        }
    },
    "allows": {
        "gh_team": "your-organization/engineering"
    }
}

What to do if a user is removed from an organization or team

A user's organization and team memberships are checked once, when they first sign in. Those teams and organizations are then persisted in the user's signed ds_actor cookie.

This means that if a user is removed from an organization or team but still has a Datasette cookie, they will still be able to access that Datasette instance.

You can remedy this by rotating the DATASETTE_SECRET environment variable any time you make changes to your GitHub organization members.

Changing this value will cause all of your existing users to be signed out, by invalidating their cookies. When they sign back in again their new memberships will be recorded in a new cookie.

See Configuring the secret in the Datasette documentation for more details.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

datasette-auth-github-0.13.1.tar.gz (6.3 kB view hashes)

Uploaded Source

Built Distribution

datasette_auth_github-0.13.1-py3-none-any.whl (10.7 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page