Building Edge Function Components
How to build, test, and push edge function components
This guide will cover how to create a new edge function component, compile it to WebAssembly, test it locally, and push it to the Edgee Component Registry.
Before diving into this guide, we highly recommend reviewing these foundational concepts and completing the CLI setup.
Edge function components are HTTP handlers powered by WebAssembly that allow you to generate custom responses such as HTML, JSON, etc. Each edge function component is configured with a specific path (or prefix) and runs in a sandboxed environment with limited access to OS functionalities.
How to create a new edge function component
Component bootstrapping
Run edgee components new
to create a new edge function component and choose your preferred programming language.
This creates a local folder containing sample code and tools to help you get started quickly.
Navigate to the new folder to begin customizing your component and implementing its business logic.
Component manifest file
Your component’s manifest file is named edgee-component.toml
. It contains all the information required to
build, test, and push your component via the Edgee CLI.
Components created via edgee components new
already include a default manifest you can easily customize.
In case you prefer managing the local folder and starting from a blank implementation, feel free to use
edgee components init
to initialize an empty manifest in the current folder.
A typical manifest file for an edge function component looks like this:
Let’s go through each section and field:
-
manifest-version
: it indicates the expected structure of your manifest file, the only available value for now is1
. -
[component]
: here you define what your component is all about, including its public name and latest versionname
: the name of this component, will be used for URLs and shown on public pages.slug
: the slug of this component, will be used for URLs and shown on public pages. Note: the slug is used as an identifier by the CLI (for a given organization). In other words, if you change a component’s slug, you will create a brand new component. A slug should change only when absolutely necessary.version
: the latest version of this component, will be used when pushing to update or create versions.language
: the programming language used this component, will be used to automate some language-specific build or compilation details. It’s automatically configured if you start with a starter template.category
: for edge function components, the value should beedge-function
.subcategory
: the available value iswasm-function
.description
: the description of this component, will be shown on public pages.documentation
: the documentation URL for this component (could also point to a public repository).repository
: the repository URL for this component (e.g. on GitHub).icon-path
: the image used on public pages to represent this component.wit-version
: the WIT interface implemented by this component. When a new WIT version is available, update this field to the latest version and the Edgee CLI will take care of updating all WIT dependencies for you.
-
[component.build]
: here you define how the component is built and where the resulting Wasm file is located.command
: the command to compile this component into Wasm, depends on your programming language of choice.output-path
: the local file path where your compiled Wasm binary is stored, will be used for local testing and pushing (make sure this is aligned with the build command defined above).
-
[component.settings.X]
: here you define as many settings as needed (see Reusability and settings below)title
: text that will be shown in the web console to identify this setting.type
: valid values arestring
,number
, orbool
.required
: (optional) a bool value to indicate whether this setting is required or not, false by default.description
: (optional) text that will be shown in the web console to describe this setting.options
: (optional) the allowed values for this setting.secret
: (optional) a bool value to indicate whether this setting value should be encrypted, false by default.
If you started via edgee components new
, most fields come with good defaults and you only need to customize
your component’s name, description, documentation URL, repository URL, and settings.
Implementing your component
An edge function component must implement the WASI HTTP interface, specifically:
This means your edge function acts as an HTTP handler that can:
- Export
wasi:http/incoming-handler@0.2.0
to handle incoming HTTP requests - Import
wasi:http/outgoing-handler@0.2.0
to make outgoing HTTP requests (if needed)
In other words, your edge function receives an HTTP request and returns an HTTP response.
The handler function automatically received all settings configured in your manifest via the
x-edgee-component-settings
HTTP header, encoded as JSON.
Don’t worry, this is already implemented in the sample code provided when using edgee components new
.
How to trigger edge functions
Edge functions are triggered by path matching using two special settings that can be configured when adding your component to an Edgee project:
edgee_path
: exact path match such as/api/auth
.edgee_path_prefix
: path prefix match such as/api/
(both/api/users
and/api/posts
will match).
When an incoming HTTP request matches your configured path, the Edgee platform will:
- Include your component settings as HTTP header
- Invoke your edge function with the corresponding HTTP request
- Return your function’s response directly to the client
Based on the programming language you chose, you’ll find the corresponding WASI HTTP bindings and unit tests to help you with the implementation.
Important Limitations: Edge functions run in a strictly sandboxed environment with limited access to OS capabilities such as file system or advanced networking. For example:
- You have no persistent storage (databases, files, or stateful data)
- You’re limited to the WASI-based HTTP interface to communicate with external services or APIs (without relying on complex SDKs or HTTP libraries)
Reusability and settings
To make your Edgee edge function component reusable for other developers and organizations via the Edgee Component Registry,
you can define a set of input settings
, containing all the configuration parameters,
API credentials, or dynamic information needed for your edge function to operate correctly.
For example, a simple API edge function might expect api_key
and allowed_origins
,
while a content serving edge function might expect content_type
and template_id
.
You can configure the expected settings for your component by adding the following for each setting in your manifest file:
Note: only title
and type
are mandatory to define a setting.
In this case, example-name
is the key your business logic will use to get its value from the settings.
Note: even if defined as number
or bool
, all settings are treated as strings when your component receives them.
For now, you’ll need to parse them into the correct type. In the future, our WIT definitions will handle types automatically as well.
Accessing settings in your function
Your edge function receives all configured settings as JSON in the x-edgee-component-settings
HTTP header.
The sample code provided when using edgee components new
already provides some utility code to parse
headers, settings, and body from the incoming HTTP request.
For example:
Build and test locally
When the implementation is ready, compile your component into Wasm:
You can customize the behavior of the build command in the manifest file by changing the target file name
and the default build script. If you’ve created a new component with edgee components new
the default build script
should be a great starting point. By default, the output of this command will be a new Wasm file in the current folder.
Note: the Edgee CLI is intended to simplify and make local development uniform across programming languages and component types.
While some of the compiling details are still visible today because of long multi-step commands or additional support scripts,
our long-term goal is to hide most of that complexity behind edgee components build
in the future.
Before pushing your component to the Edgee Component Registry, it’s highly recommended to validate and test the Wasm file locally:
The test command starts a local HTTP server running your edge function. You can use curl, your browser, or another HTTP client to interact with the local server and test your component’s behavior.
Note: the test command is using your Wasm file, not your source code directly. So don’t forget to re-compile your component to Wasm after code updates.
Testing with local HTTP requests
For example, a test HTTP request might look like:
Your edge function will receive this as a WASI HTTP request object and can:
- Inspect request headers and body
- Extract encoded data from the request
- Make HTTP requests to external services to send or fetch data
- Generate and return a custom HTTP response such as HTML or JSON
Note: you can craft custom test requests that better suit your component’s needs by creating test scenarios in your component’s test suite.
Verify HTTP responses
By default, the test command runs your local Wasm file and shows you:
- The incoming HTTP request your edge function received
- Any outgoing HTTP requests your edge function made
- The final HTTP response your edge function returned
Once the test server is running, you can test your edge function with curl:
The test server will pass your configured settings to the edge function via the x-edgee-component-settings
header.
Note: if your component interacts with external services, we recommend using test or staging environments whenever possible to avoid affecting production systems or external APIs during testing.
Push to the registry
When your component is ready for deployment, it’s time to push it to the registry.
Make sure you went through these steps so far:
- Installed and configured the Edgee CLI
- Implemented the WASI HTTP interface
- Updated your manifest file with the correct build command and settings
- Validated the Wasm file with local tests
- Decided on the component’s visibility (public or private)
Then run the following command:
Your new component will be pushed to the Edgee Component Registry under your Organization.
In case your Edgee user is part of multiple Organizations, you can provide the organization identifier
via edgee components push ORG_ID
.
Congratulations! Your edge function component is now available on the Edgee Component Registry at https://www.edgee.cloud/{organization}/{component}
.
In case you pushed a private component, its page is only visible to you.
You can also view and edit it at https://www.edgee.cloud/~/registry/{organization}
.
Publish/unpublish components
A public component is visible to anyone in the registry, while private components are only available to a specific organization.
During the initial push, you decide between private or public. Every subsequent push command will keep the same visibility. If you need to publish a private component or unpublish a public component after the initial push, you’ll need to declare the visibility change explicitly:
Note: unpublishing a public component is only possible as long as the component isn’t used by any project. You can still push the same component as private for a specific organization. This might be useful for special use cases where a component needs ad-hoc customizations.
Archive/delete components
After pushing, visit https://www.edgee.cloud/~/registry/{organization}
and select your component to edit it.
Here you can archive or delete it with the corresponding buttons.
Please note that:
- Archived components become invisible on public pages, but can be unarchived. Existing projects using an archived component still work fine.
- Deleted components simply stop existing and this action is irreversible. If there is at least one Edgee project using a component, you cannot delete it.
Note: even though the web console allows you to edit most fields and create new versions, we recommend using the Edgee CLI to push updates and new versions. This way, your manifest file remains the source of truth.
Use the edge function component
Once your edge function component is available via the Edgee Component Registry, you can add it to a project and it will start processing HTTP requests.
First of all, you’ll need to create a project and configure it for your website. Unlike data collection components, edge functions don’t require the Edgee SDK - they work directly with HTTP traffic.
Visit the Components section of your project and click Add a component. Here you can filter by category, organization, and visibility. Make sure to select Private in the bottom-left if you pushed a private component. Select edge-function in the category filter to see only edge function components.
Select your edge function component and configure it. A few things to keep in mind:
- You can select a specific version of your component, by default the latest version is selected.
- Enable Auto update if you want to automatically use the latest version of your component for this project.
- Configure the path matching by setting either:
edgee_path
: For exact path matches (e.g.,/api/auth
)edgee_path_prefix
: For path prefix matches (e.g.,/api/
matches all paths starting with/api/
)
Once the project is set up and your edge function component is active, it will start processing HTTP requests that match your configured path.
Your edge function can:
- Generate custom API responses
- Serve static content or dynamic responses
- Implement simple authentication endpoints
- Create redirect responses
- Make HTTP requests to external services to send or fetch data
- Return responses with custom headers and status codes
Performance Considerations: Edge functions run on every matching request in a WebAssembly sandbox. Keep implementations lightweight and efficient since they cannot access persistent storage or maintain state.