An SDK to easily create Farcaster Frames, manage frame analytics and pin images to IPFS using Pinata.
View the full documentation here.
npm i pinata-fdk
yarn add pinata-fdk
If you want to leverage IPFS pinning capabilities, you must enter your Pinata JWT and a Pinata gateway during intialization.
import { PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: "<YOUR_PINATA_JWT>",
pinata_gateway: "<YOUR_PINATA_GATEWAY>"},
);
If you are only using the frame metadata functionality, you do not need to enter your credentials.
import { PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK();
The FDK makes it easy to give your Farcaster app write access for users so you can do things like sending casts or following other users. To get a better concept of the flow of Farcaster Auth, check out the guide here.
This function will create a signer with Farcaster Auth, sign the key with your Farcaster App mnemonic phrase and FID, then send a request to Warpcast to register the signer. For more info on those please see these docs.
In order to use it make sure the mnemonic and FID are included with the PinataFDK instance.
import { PinataFDK } from "pinata-fdk"
const fdk = new PinataFDK({
pinata_jwt: 'YOUR_PINATA_JWT',
pinata_gateway: 'YOUR_GATEWAY',
app_fid: 'APP_FID',
app_mnemonic: 'APP_MNEMONIC'
})
import { PinataFDK } from "pinata-fdk"
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
app_fid: `${process.env.APP_FID}`
app_mnemonic: `${process.env.FARCASTER_DEVELOPER_MNEMONIC}`
})
const signerData: WarpcastPayload = await fdk.createSigner()
After creating the signer the user would visit the signerData.deep_link_url
which would open Warpcast on their account to approve the signer. The user will have to pay warps since it is not sponsored.
{
signer_id: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
token: "0xe3bffad26b16cf825f3d062d",
deep_link_url: "farcaster://signed-key-request?token=0xe3bffad26b16cf825f3d062d",
status: "pending_approval"
}
Sponsored signers is very similar to createSigner
except the end user will not have to pay warps to approve the signer. However there may be limits to how many sponsored signers you can use based on your Pinata plan. You still need the app_mnemonic
to sign the key being used.
import { PinataFDK } from "pinata-fdk"
const fdk = new PinataFDK({
pinata_jwt: 'YOUR_PINATA_JWT',
pinata_gateway: 'YOUR_GATEWAY',
app_fid: 'APP_FID',
app_mnemonic: 'APP_MNEMONIC'
})
import { PinataFDK } from "pinata-fdk"
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
app_fid: `${process.env.APP_FID}`,
app_mnemonic: `${process.env.FARCASTER_DEVELOPER_MNEMONIC}`
})
const signerData: WarpcastPayload = await fdk.createSponsoredSigner()
After creating the signer the user would visit the signerData.deep_link_url
which would open Warpcast on their account to approve the signer. Since it is sponsored the user will not have to pay warps to sign in, however it will show Pinata as the app.
{
signer_id: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
token: "0x21658c8fa560aca0f35a5e4a",
deep_link_url: "farcaster://signed-key-request?token=0x21658c8fa560aca0f35a5e4a",
status: "pending_approval"
}
After creating a signer and giving the user the deep_link_url
you will want to poll the signer to see if they have approved it and record the response to your account.
pollSigner
takes a parameter of token
which is provided in either createSigner
or createSponsoredSigner
.
import { PinataFDK } from "pinata-fdk"
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
app_fid: `${process.env.APP_FID}`
})
const pollData = await fdk.pollSigner("0x21658c8fa560aca0f35a5e4a")
{
token: "0x321bbb927d9009232a7c26d6",
deeplinkUrl: "farcaster://signed-key-request?token=0x321bbb927d9009232a7c26d6",
key: "0x858e9ed1af97ec0c1cf06e7d769a2bca9ec324c152f320ee34b253af27b486f4",
requestFid: 20918,
state: "pending",
isSponsored: false
}
After you have created a signer and it has been polled as complete, you can fetch the signer at any time using getSigner
- fid - A number/integer of the FID you want to query. (Optional)
import { PinataFDK } from "pinata-fdk"
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
app_fid: `${process.env.APP_FID}`
})
const pollData = await fdk.getSigner(6023)
{
signers: [
{
id: 57,
signer_uuid: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
fid: 6023,
public_key: "dad973170c63739f7c812d188fab1df074eb1cd48facf6556e2ef9cbb76b4c18",
signer_approved: true,
revoked: false
}
],
next_page_token: "eyJvZmZzZXQiOiI1NyJ9"
}
These methods can be used to create casts, react to them, or even follow users. All of them require a signerId
which is a result of using Farcaster Auth, as well as the PinataFDK initialization with a Pinata JWT.
All of these methods also return the same response, which is the direct response from the Pinata Hub.
type CastResponse = {
data: {
type: string;
fid: number;
timestamp: number;
network: string;
castAddBody: {
embedsDeprecated: any[];
mentions: any[];
text: string;
mentionsPositions: any[];
embeds: any[];
};
};
hash: string;
hashScheme: string;
signature: string;
signatureScheme: string;
signer: string;
dataBytes: string;
};
With the sendCast
method you can effortlessly post to Farcaster using a signerId
.
- CastRequest - An object that contains the
signerId
and thecastAddBody
which follows the Farcaster Hub standard for sending casts.
type CastRequest = {
signerId: string;
castAddBody: CastBody;
}
type CastBody = {
embedsDeprecated?: string[];
mentions?: number[];
parentCastId?: CastId | null;
parentUrl?: string | null;
text?: string | null;
mentionsPositions?: number[] | null;
embeds?: Embed[] | null;
}
type CastId = {
fid: number;
hash: string;
}
type Embed = {
url?: string | null;
castId?: CastId | null;
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const res: CastResponse = await fdk.sendCast({
castAddBody: {
text: "Hello World from !",
mentions: [6023],
mentionsPositions: [18],
parentUrl: "https://warpcast.com/~/channel/pinata",
embeds: [
{
url: "https://pinata.cloud"
},
{
castId: {
fid: 6023
hash: "0xcae8abd9badbb60c9b610ec264f42ed9f1785c6f",
}
}
],
parentCastId: {
fid: 6023,
hash: "0xcae8abd9badbb60c9b610ec264f42ed9f1785c6f"
}
},
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1"
});
This method can delete a cast with a provided target hash.
- hash - Target hash of the cast that needs to be deleted
- signerId - Signer for the cast
type CastDelete = {
hash: string;
signerId: string;
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const deleteReq: CastResponse = await fdk.deleteCast({
hash: "0x490889854a4f3233433b1ad0560f016f04feeeff",
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
This method can like a cast based on the provided target hash.
- hash - Hash of the target cast to be liked
- signerId - Signer of the user liking the cast
type LikeCast = {
hash: string,
signerId: string
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const likeReq: CastResponse = await fdk.likeCast({
hash: "0x490889854a4f3233433b1ad0560f016f04feeeff",
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
If a cast is already liked by the user, this method will unlike it.
- hash - Hash of the target cast to be unliked
- signerId - Signer of the user unliking the cast
type LikeCast = {
hash: string,
signerId: string
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const unlikeReq: CastResponse = await fdk.unlikeCast({
hash: "0x490889854a4f3233433b1ad0560f016f04feeeff",
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
This method will recast a cast based on the target hash.
- hash - Hash of the target cast to be recast
- signerId - Signer of the user recasting target cast
type RecastCast = {
hash: string;
signerId: string;
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const recastCastReq: CastResponse = await fdk.recastCast({
hash: "0x490889854a4f3233433b1ad0560f016f04feeeff",
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
Works just like recastCast
but removes an existing recast based on a hash of the target cast.
- hash - Hash of the target cast to remove recast
- signerId - Signer of the user removing the recast
type RecastCast = {
hash: string;
signerId: string;
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const removeRecastReq: CastResponse = await fdk.removeRecast({
hash: "0x490889854a4f3233433b1ad0560f016f04feeeff",
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
Follows a target user based on their FID.
- fid - The FID of the target user to follow
- signerId - The signer for the user following the target user
type FollowUser = {
fid: number;
signerId: string;
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const followUserReq: CastResponse = await fdk.followUser({
fid: 6023,
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
Unfollows a user that the signer is already following.
- fid - The FID of the target user to unfollow
- signerId - The signer for the user unfollowing the target user
type FollowUser = {
fid: number;
signerId: string;
}
import { CastResponse, PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: `${process.env.PINATA_JWT}`,
pinata_gateway: "",
});
const unfollowUserReq: CastResponse = await fdk.unfollowUser({
fid: 6023,
signerId: "ba2d9f6d-7514-4967-8b52-5a040b7da4a1",
};
You can use this function to easily create the Farcaster specific metadata needed for your frame.
The only required input is cid
or image
.
buttons
- An array of button specifications (max 4). (Optional)image
- A string for a valid hosted image url.cid
- A string representing the cid of an IPFS pinned image.input
- A string representing the text displayed for text input. (Optional)post_url
- A string which contains a valid URL to send the Signature Packet to. (Optional)refresh_period
- A string representing the refresh period for the image used. (Optional)aspect_ratio
- A string representing the aspect ratio for the image used. (Optional)state
- An object (e.g. JSON) representing the state data for the frame. (Optional)
Note: state should only be included in response frames, not initial frames.
type FrameHTMLType = {
buttons?: [FrameButtonMetadata, ...FrameButtonMetadata[]];
image?: {url: string};
cid?: string;
input?: FrameInputMetadata;
post_url?: string;
refresh_period?: number;
aspect_ratio?: "1.91:1" | "1:1",
state?: object
} & (
{ image: {url: string}}| { cid: string }
);
type FrameButtonMetadata = {
label: string;
action?: "post" | "post_redirect" | "mint" | "link" | "tx";
target?: string;
}
type FrameInputMetadata = {
text: string;
};
const frameMetadata = fdk.getFrameMetadata({
post_url: `<YOUR_DOMAIN>/api/test`,
input: {text: "Hello, world!"},
aspectRatio: "1.91:1",
buttons: [
{ label: 'Click me', action: 'post'},
{ label: 'Button 2', action: "post_redirect"},
{ label: 'Button 3', action: "mint" },
{ label: 'Button 4', action: "link" },
],
cid: "<YOUR_CID>",
state: {counter: 1}
});
<meta name="fc:frame" content="vNext">
<meta name="og:image" content="https://<YOUR_GATEWAY>/ipfs/<YOUR_CID>">
<meta name="fc:frame:image" content="https://<YOUR_GATEWAY>/ipfs/<YOUR_CID>">
<meta name="fc:frame:aspectRatio" content="1.91:1">
<meta name="fc:frame:input:text" content="Hello, world!">
<meta name="fc:frame:button:1" content="Click me">
<meta name="fc:frame:button:1:action" content="post">
<meta name="fc:frame:button:2" content="Button 2">
<meta name="fc:frame:button:2:action" content="post_redirect">
<meta name="fc:frame:button:3" content="Button 3">
<meta name="fc:frame:button:3:action" content="mint">
<meta name="fc:frame:button:4" content="Button 4">
<meta name="fc:frame:button:4:action" content="link">
<meta name="fc:frame:post_url" content="<YOUR_DOMAIN>/api/test">
<meta name="fc:frame:state" content="%7B%22counter%22%3A1%7D">
There are two different ways to set the images of your frame metadata.
⚡️ Raw URL
⚡️ CID
Specify a hosted url image link.
const frameMetadata = fdk.getFrameMetadata({
image: { url: "<YOUR_URL>"}
});
Specify a CID from your Pinata account.
const frameMetadata = fdk.getFrameMetadata({
cid: "QmX63EYiDk9cExrv4GDmZ5soJKkgqoUJv9LbtPyugLBtV2"
});
//Must insert Pinata credentials when intializing SDK.
Uploads an image to IPFS from a url. This url may be passed to the getFrameMetadata
function.
url
- A string for a valid hosted url image.
const ipfsUrl = await fdk.convertUrlToIPFS("https://example.com");
const frameMetadata = fdk.getFrameMetadata({
image: { url: ipfsUrl}
});
https://<YOUR_GATEWAY>/ipfs/<YOUR_CID>
Returns a Promise that indicates wether a message signature is valid by querying Pinata's Farcaster hub.
body
- An object representing the raw payload of an action frame produced by Farcaster.
{
untrustedData: {
fid: 2,
url: "https://fcpolls.com/polls/1",
messageHash: "0xd2b1ddc6c88e865a33cb1a565e0058d757042974",
timestamp: 1706243218,
network: 1,
buttonIndex: 2,
inputText: "hello world", // "" if requested and no input, undefined if input not requested
castId: {
fid: 226,
hash: "0xa48dd46161d8e57725f5e26e34ec19c13ff7f3b9",
},
},
trustedData: {
messageBytes: "d2b1ddc6c88e865a33cb1a565e0058d757042974...",
},
};
const { isValid, message } = await fdk.validateFrameMessage(body);
{
isValid: true,
message: {
data: {
type: 13,
fid: 15974,
timestamp: 98469569,
network: 1,
castAddBody: undefined,
castRemoveBody: undefined,
reactionBody: undefined,
verificationAddAddressBody: undefined,
verificationRemoveBody: undefined,
userDataBody: undefined,
linkBody: undefined,
usernameProofBody: undefined,
frameActionBody: [Object]
},
hash: <Buffer 46 f2 c3 eb ee 56 92 4f 14 47 2e f0 22 de 41 b6 26 52 a8 4d>,
hashScheme: 1,
signature: <Buffer 7a a0 5e b1 5a f9 59 a7 08 e9 d0 19 d2 24 47 f7 4b 15 d8 70 a8 fb 7e 36 f1 b6 06 14 c2 63 db 7f eb 98 4c 8a e7 98 5c 0d 72 27 04 f3 e9 19 08 10 e7 e9 ... 14 more bytes>,
signatureScheme: 1,
signer: <Buffer 1d 46 b7 63 1c 6f 15 7b 86 3a c1 9a 64 cd ba 7e b1 cc 2a cb 81 53 37 99 e8 69 e1 49 11 c4 81 4e>,
dataBytes: undefined
}
}
Returns the connected Ethereum address for an FID.
fid
- A number representing the fid of the user.
const address = await fdk.getAddressForFid(15974);
"0x9b7c18a71a98acd2f1271e2d1fe63750a70bc52b"
Returns the user datadata for an FID.
fid
- A number representing the fid of the user.
const userData = await fdk.getUserByFid(20591);
{
fid: 20591,
username: 'kyletut',
pfp: 'https://i.imgur.com/TLMFnH6.jpg',
bio: 'Everyone is from somewhere. Cofounder and CEO of Pinata. https://www.pinata.cloud/farcaster'
}
To get started visit the Integrations page by clicking on the profile button in the top right, then selecting Integrations.
Note: You must have Warpcast integrated in the web app to use this functionality.
In any of your POST endpoints for frames, you can send analytics like this:*
Sends data to Pinata analytics for a specific frame.
frame_data
- An object representing the raw payload of an action frame produced by Farcaster.frame_id
- A string representing the frame you want to track.custom_id
: A string representing a unique identifier to segment requests within the specified frame. (Optional)
// This should be the raw payload from the frame action
const frame_data = {
untrustedData: {
fid: 2,
url: "https://fcpolls.com/polls/1",
messageHash: "0xd2b1ddc6c88e865a33cb1a565e0058d757042974",
timestamp: 1706243218,
network: 1,
buttonIndex: 2,
inputText: "hello world", // "" if requested and no input, undefined if input not requested
castId: {
fid: 226,
hash: "0xa48dd46161d8e57725f5e26e34ec19c13ff7f3b9",
},
},
trustedData: {
messageBytes: "d2b1ddc6c88e865a33cb1a565e0058d757042974...",
},
};
const frame_id = "my-unique-frame-name"
const custom_id = "my_custom_id"
await fdk.sendAnalytics(frame_id, frame_data, custom_id)
{success: true}
After this is deployed you will see the analytics on the Frame Analytics Page.
If you are using the Frog framework, you can utilize Pinata Frame analytics by importing our custom middleware function analyticsMiddleware
.
import { PinataFDK } from "pinata-fdk";
const fdk = new PinataFDK({
pinata_jwt: "<YOUR_PINATA_JWT>",
pinata_gateway: "<YOUR_PINATA_GATEWAY>"},
);
const app = new Frog({
basePath: '/api',
// hubApiUrl: "https://hub.pinata.cloud"
})
app.use("/", fdk.analyticsMiddleware({ frameId: "frame_id", customId: "custom_id"}));
Pin files directly from your src
folder using the pinata-fdk npx pin
command.
Create a pins
folder located under your src
folder
src
-pins/
Add any images you want uploaded to IPFS!
src
-pins/
--image.png
Run the command:
npx pin <YOUR_PINATA_JWT>
Check your src/pins
folder for a pins_list.json
file. Here you can easily access your CID to input into getFrameMetadata
.