Solana Actions and Blockchain Links (Blinks)
Read the docs to get started
Solana Actions are specification-compliant APIs that return transactions on the
Solana blockchain to be previewed, signed, and sent across a number of various
contexts, including QR codes, buttons + widgets, and websites across the
internet. Actions make it simple for developers to integrate the things you can
do throughout the Solana ecosystem right into your environment, allowing you to
perform blockchain transactions without needing to navigate away to a different
app or webpage.
Blockchain links – or blinks – turn any Solana Action into a shareable,
metadata-rich link. Blinks allow Action-aware clients (browser extension
wallets, bots) to display additional capabilities for the user. On a website, a
blink might immediately trigger a transaction preview in a wallet without going
to a decentralized app; in Discord, a bot might expand the blink into an
interactive set of buttons. This pushes the ability to interact on-chain to any
web surface capable of displaying a URL.
Simplified Type Definitions
The types and interfaces declared within this readme files are often the
simplified version of the types to aid in readability.
For better type safety and improved developer experience, the
@solana/actions-spec
package contains more complex type definitions. You can
find the
source code for them here.
Contributing
If you would like to propose an update the Solana Actions specification, please
publish a proposal on the official Solana forum under the Solana Request for
Comments (sRFC) section: https://forum.solana.com/c/srfc/6
Specification
The Solana Actions specification consists of key sections that are part of a
request/response interaction flow:
Each of these requests are made by the Action client (e.g. wallet app, browser
extension, dApp, website, etc) to gather specific metadata for rich user
interfaces and to facilitate user input to the Actions API.
Each of the responses are crafted by an application (e.g. website, server
backend, etc) and returned to the Action client. Ultimately, providing a
signable transaction or message for a wallet to prompt the user to approve,
sign, and send to the blockchain.
URL Scheme
A Solana Action URL describes an interactive request for a signable Solana
transaction or message using the solana-action
protocol.
The request is interactive because the parameters in the URL are used by a
client to make a series of standardized HTTP requests to compose a signable
transaction or message for the user to sign with their wallet.
solana-action:<link>
-
A single link
field is required as the pathname. The value must be a
conditionally
URL-encoded
absolute HTTPS URL.
-
If the URL contains query parameters, it must be URL-encoded. URL-encoding the
value prevents conflicting with any Actions protocol parameters, which may be
added via the protocol specification.
-
If the URL does not contain query parameters, it should not be URL-encoded.
This produces a shorter URL and a less dense QR code.
In either case, clients must
URL-decode
the value. This has no effect if the value isn't URL-encoded. If the decoded
value is not an absolute HTTPS URL, the wallet must reject it as malformed.
OPTIONS response
In order to allow Cross-Origin Resource Sharing
(CORS) within Actions
clients (including blinks), all Action endpoints should respond to HTTP requests
for the OPTIONS
method with valid headers that will allow clients to pass CORS
checks for all subsequent requests from their same origin domain.
An Actions client may perform
"preflight"
requests to the Action URL endpoint in order check if the subsequent GET request
to the Action URL will pass all CORS checks. These CORS preflight checks are
made using the OPTIONS
HTTP method and should respond with all required HTTP
headers that will allow Action clients (like blinks) to properly make all
subsequent requests from their origin domain.
At a minimum, the required HTTP headers include:
Access-Control-Allow-Origin
with a value of *
- this ensures all Action clients can safely pass CORS checks in order to make
all required requests
Access-Control-Allow-Methods
with a value of GET,POST,PUT,OPTIONS
- ensures all required HTTP request methods are supported for Actions
Access-Control-Allow-Headers
with a minimum value of
Content-Type, Authorization, Content-Encoding, Accept-Encoding
For simplicity, developers should consider returning the same response and
headers to OPTIONS
requests as their GET
response.
The actions.json
file response must also return valid Cross-Origin headers
for GET
and OPTIONS
requests, specifically the
Access-Control-Allow-Origin
header value of *
.
See actions.json below for more details.
GET Request
The Action client (e.g. wallet, browser extension, etc) should make an HTTP
GET
JSON request to the Action's URL endpoint.
- The request should not identify the wallet or the user.
- The client should make the request with an
Accept-Encoding
header. - The client should display the domain of the URL as the request is being made.
GET Response
The Action's URL endpoint (e.g. application or server backend) should respond
with an HTTP OK
JSON response (with a valid payload in the body) or an
appropriate HTTP error.
Error responses (i.e. HTTP 4xx and 5xx status codes) should return a JSON
response body following ActionError
to present a helpful error message to
users. See Action Errors.
GET Response Body
A GET
response with an HTTP OK
JSON response should include a body payload
that follows the interface specification:
export type ActionType = "action" | "completed";
export type ActionGetResponse = Action<"action">;
export interface Action<T extends ActionType = "action"> {
type: T;
icon: string;
title: string;
description: string;
label: string;
disabled?: boolean;
links?: {
actions: LinkedAction[];
};
error?: ActionError;
}
-
type
- The type of action being given to the user. Defaults to action
. The
initial ActionGetResponse
is required to have a type of action
.
action
- Standard action that will allow the user to interact with any of
the LinkedActions
completed
- Used to declare the "completed" state within action chaining.
-
icon
- The value must be an absolute HTTP or HTTPS URL of an icon image. The
file must be an SVG, PNG, or WebP image, or the client/wallet must reject it
as malformed.
-
title
- The value must be a UTF-8 string that represents the source of the
action request. For example, this might be the name of a brand, store,
application, or person making the request.
-
description
- The value must be a UTF-8 string that provides information on
the action. The description should be displayed to the user.
-
label
- The value must be a UTF-8 string that will be rendered on a button
for the user to click. All labels should not exceed 5 word phrases and should
start with a verb to solidify the action you want the user to take. For
example, "Mint NFT", "Vote Yes", or "Stake 1 SOL".
-
disabled
- The value must be boolean to represent the disabled state of the
rendered button (which displays the label
string). If no value is provided,
disabled
should default to false
(i.e. enabled by default). For example,
if the action endpoint is for a governance vote that has closed, set
disabled=true
and the label
could be "Vote Closed".
-
error
- An optional error indication for non-fatal errors. If present, the
client should display it to the user. If set, it should not prevent the client
from interpreting the action or displaying it to the user (see
Action Errors). For example, the error can be used together
with disabled
to display a reason like business constraints, authorization,
the state, or an error of external resource.
-
links.actions
- An optional array of related actions for the endpoint. Users
should be displayed UI for each of the listed actions and expected to only
perform one. For example, a governance vote action endpoint may return three
options for the user: "Vote Yes", "Vote No", and "Abstain from Vote".
-
If no links.actions
is provided, the client should render a single button
using the root label
string and make the POST request to the same action
URL endpoint as the initial GET request.
-
If any links.actions
are provided, the client should only render buttons
and input fields based on the items listed in the links.actions
field. The
client should not render a button for the contents of the root label
.
export interface LinkedAction {
href: string;
label: string;
parameters?: Array<TypedActionParameter>;
}
The ActionParameter
allows declaring what input the Action API is requesting
from the user:
export interface ActionParameter {
type?: ActionParameterType;
name: string;
label?: string;
required?: boolean;
pattern?: string;
patternDescription?: string;
min?: string | number;
max?: string | number;
}
The pattern
should be a string equivalent of a valid regular expression. This
regular expression pattern should by used by blink-clients to validate user
input before before making the POST request. If the pattern
is not a valid
regular expression, it should be ignored by clients.
The patternDescription
is a human readable description of the expected input
requests from the user. If pattern
is provided, the patternDescription
is
required to be provided.
The min
and max
values allows the input to set a lower and/or upper bounds
of the input requested from the user (i.e. min/max number and or min/max
character length), and should be used for client side validation. For input
type
s of date
or datetime-local
, these values should be a string dates.
For other string based input type
s, the values should be numbers representing
their min/max character length.
If the user input value is not considered valid per the pattern
, the user
should receive a client side error message indicating the input field is not
valid and displayed the patternDescription
string.
The type
field allows the Action API to declare more specific user input
fields, providing better client side validation and improving the user
experience. In many cases, this type will resemble the standard
HTML input element.
The ActionParameterType
can be simplified to the following type:
export type ActionParameterType =
| "text"
| "email"
| "url"
| "number"
| "date"
| "datetime-local"
| "checkbox"
| "radio"
| "textarea"
| "select";
Each of the type
values should normally result in a user input field that
resembles a standard HTML input
element of the corresponding type
(i.e.
<input type="email" />
) to provide better client side validation and user
experience:
text
- equivalent of HTML
“text” input
elementemail
- equivalent of HTML
“email” input
elementurl
- equivalent of HTML
“url” input
elementnumber
- equivalent of HTML
“number” input
elementdate
- equivalent of HTML
“date” input
elementdatetime-local
- equivalent of HTML
“datetime-local” input
elementcheckbox
- equivalent to a grouping of standard HTML
“checkbox” input
elements. The Action API should return options
as detailed below. The user
should be able to select multiple of the provided checkbox options.radio
- equivalent to a grouping of standard HTML
“radio” input
elements. The Action API should return options
as detailed below. The user
should be able to select only one of the provided radio options.- Other HTML input type equivalents not specified above (
hidden
, button
,
submit
, file
, etc) are not supported at this time.
In addition to the elements resembling HTML input types above, the following
user input elements are also supported:
textarea
- equivalent of HTML
textarea element.
Allowing the user provide multi-line input.select
- equivalent of HTML
select element,
allowing the user to experience a “dropdown” style field. The Action API
should return options
as detailed below.
When type
is set as select
, checkbox
, or radio
then the Action API
should include an array of options
that each provide a label
and value
at
a minimum. Each option may also have a selected
value to inform the
blink-client which of the options should be selected by default for the user
(see checkbox
and radio
for differences).
This ActionParameterSelectable
can be simplified to the following type
definition:
interface ActionParameterSelectable extends ActionParameter {
options: Array<{
label: string;
value: string;
selected?: boolean;
}>;
}
If no type
is set or an unknown/unsupported value is set, blink-client should
default to text
and render a simple text input.
The Action API is still responsible to validate and sanitize all data from the
user input parameters, enforcing any “required” user input as necessary.
For platforms other that HTML/web based ones (like native mobile), the
equivalent native user input component should be used to achieve the equivalent
experience and client side validation as the HTML/web input types described
above.
POST Request
The client must make an HTTP POST
JSON request to the action URL with a body
payload of:
{
"account": "<account>"
}
account
- The value must be the base58-encoded public key of an account that
may sign the transaction.
The client should make the request with an
Accept-Encoding header
and the application may respond with a
Content-Encoding header
for HTTP compression.
The client should display the domain of the action URL as the request is being
made. If a GET
request was made, the client should also display the title
and render the icon
image from that GET response.
POST Response
The Action's POST
endpoint should respond with an HTTP OK
JSON response
(with a valid payload in the body) or an appropriate HTTP error.
Error responses (i.e. HTTP 4xx and 5xx status codes) should return a JSON
response body following ActionError
to present a helpful error message to
users. See Action Errors.
POST Response Body
A POST
response with an HTTP OK
JSON response should include a body payload
of:
export interface ActionPostResponse<T extends ActionType = ActionType> {
transaction: string;
message?: string;
links?: {
next: NextActionLink;
};
}
-
transaction
- The value must be a base64-encoded
serialized transaction.
The client must base64-decode the transaction and
deserialize it.
-
message
- The value must be a UTF-8 string that describes the nature of the
transaction included in the response. The client should display this value to
the user. For example, this might be the name of an item being purchased, a
discount applied to a purchase, or a thank you note.
-
links.next
- An optional value use to "chain" multiple Actions together in
series. After the included transaction
has been confirmed on-chain, the
client can fetch and render the next action. See
Action Chaining for more details.
-
The client and application should allow additional fields in the request body
and response body, which may be added by future specification updates.
The application may respond with a partially or fully signed transaction. The
client and wallet must validate the transaction as untrusted.
POST Response - Transaction
If the transaction
signatures
are empty or the transaction has NOT been partially signed:
- The client must ignore the
feePayer
in the transaction and set the feePayer
to the account
in the request. - The client must ignore the
recentBlockhash
in the transaction and set the recentBlockhash
to the
latest blockhash. - The client must serialize and deserialize the transaction before signing it.
This ensures consistent ordering of the account keys, as a workaround for
this issue.
If the transaction has been partially signed:
- The client must NOT alter the
feePayer
or
recentBlockhash
as this would invalidate any existing signatures. - The client must verify existing signatures, and if any are invalid, the client
must reject the transaction as malformed.
The client must only sign the transaction with the account
in the request, and
must do so only if a signature for the account
in the request is expected.
If any signature except a signature for the account
in the request is
expected, the client must reject the transaction as malicious.
Action Chaining
Solana Actions can be "chained" together in a successive series. After an
Action's transaction is confirmed on-chain, the next action can be obtained and
presented to the user.
Action chaining allows developers to build more complex and dynamic experiences
within blinks, including:
- providing multiple transactions (and eventually sign message) to a user
- customized action metadata based on the user's wallet address
- refreshing the blink metadata after a successful transaction
- receive an API callback with the transaction signature for additional
validation and logic on the Action API server
- customized "success" messages by updating the displayed metadata (e.g. a new
image and description)
To chain multiple actions together, in any ActionPostResponse
include a
links.next
of either:
PostNextActionLink
- POST request link with a same origin callback url to
receive the signature
and user's account
in the body. This callback url
should respond with a NextAction
.InlineNextActionLink
- Inline metadata for the next action to be presented
to the user immediately after the transaction has confirmed. No callback will
be made.
export type NextActionLink = PostNextActionLink | InlineNextActionLink;
export interface PostNextActionLink {
type: "post";
href: string;
}
export interface InlineNextActionLink {
type: "inline";
action: NextAction;
}
NextAction
After the ActionPostResponse
included transaction
is signed by the user and
confirmed on-chain, the blink client should either:
- execute the callback request to fetch and display the
NextAction
, or - if a
NextAction
is already provided via links.next
, the blink client
should update the displayed metadata and make no callback request
If the callback url is not the same origin as the initial POST request, no
callback request should be made. Blink clients should display an error notifying
the user.
export type NextAction = Action<"action"> | CompletedAction;
export type CompletedAction = Omit<Action<"completed">, "links">;
Based on the type
, the next action should be presented to the user via blink
clients in one of the following ways:
-
action
- (default) A standard action that will allow the user to see the
included Action metadata, interact with the provided LinkedActions
, and
continue to chain any following actions.
-
completed
- The terminal state of an action chain that can update the blink
UI with the included Action metadata, but will not allow the user to execute
further actions.
If no links.next
is not provided, blink clients should assume the current
action is final action in the chain, presenting their "completed" UI state after
the transaction is confirmed.
Action Errors
Actions APIs should return errors using ActionError
in order to present
helpful error messages to the user. Depending on the context, this error could
be fatal or non-fatal.
export interface ActionError {
message: string;
}
When an Actions API responds with an HTTP error status code (i.e. 4xx and 5xx),
the response body should be a JSON payload following ActionError
. The error is
considered fatal and the included message
should be presented to the user.
For API responses that support the optional error
attribute (like
ActionGetResponse
), the error is considered non-fatal and the
included message
should be presented to the user.
actions.json
The purpose of the actions.json
file allows an application to
instruct clients on what website URLs support Solana Actions and provide a
mapping that can be used to perform GET requests to an Actions
API server.
The actions.json
file response must also return valid Cross-Origin headers
for GET
and OPTIONS
requests, specifically the
Access-Control-Allow-Origin
header value of *
.
See OPTIONS response above for more details.
The actions.json
file should be stored and universally accessible at the root
of the domain.
For example, if your web application is deployed to my-site.com
then the
actions.json
file should be accessible at https://my-site.com/actions.json
.
This file should also be Cross-Origin accessible via any browser by having a
Access-Control-Allow-Origin
header value of *
.
Rules
The rules
field allows the application to map a set of a website's relative
route paths to a set of other paths.
Type: Array
of ActionRuleObject
.
interface ActionRuleObject {
pathPattern: string;
apiPath: string;
}
Rules - pathPattern
A pattern that matches each incoming pathname. It can be an absolute or relative
path and supports the following formats:
-
Exact Match: Matches the exact URL path.
- Example:
/exact-path
- Example:
https://website.com/exact-path
-
Wildcard Match: Uses wildcards to match any sequence of characters in the
URL path. This can match single (using *
) or multiple segments (using **
).
(see Path Matching below).
- Example:
/trade/*
will match /trade/123
and /trade/abc
, capturing only
the first segment after /trade/
. - Example:
/category/*/item/**
will match /category/123/item/456
and
/category/abc/item/def
. - Example:
/api/actions/trade/*/confirm
will match
/api/actions/trade/123/confirm
.
Rules - apiPath
The destination path for the action request. It can be defined as an absolute
pathname or an external URL.
- Example:
/api/exact-path
- Example:
https://api.example.com/v1/donate/*
- Example:
/api/category/*/item/*
- Example:
/api/swap/**
Rules - Query Parameters
Query parameters from the original URL are always preserved and appended to the
mapped URL.
Rules - Path Matching
The following table outlines the syntax for path matching patterns:
Operator | Matches |
---|
* | A single path segment, not including the surrounding path separator / characters. |
** | Matches zero or more characters, including any path separator / characters between multiple path segments. If other operators are included, the ** operator must be the last operator. |
? | Unsupported pattern. |
License
The Solana Actions JavaScript SDK is open source and available under the Apache
License, Version 2.0. See the LICENSE file for more info.