Goldsky

https://docs.goldsky.com
Admin
Goldsky is a modern backend platform for crypto-enabled products that provides instant GraphQL APIs,...
Tokens:626,305
Snippets:4,620
Trust Score:8.9
Update:1 day ago
Show doc for...
Tokens:
Raw
### Clone and install the example repository

Source: https://docs.goldsky.com/compose/guides/build-a-prediction-market

Commands to clone the documentation-examples repository, navigate to the prediction-market directory, and install dependencies.

```bash
git clone https://github.com/goldsky-io/documentation-examples.git
cd documentation-examples/compose/prediction-market
npm install
```

--------------------------------

### Start with small resource size

Source: https://docs.goldsky.com/turbo-pipelines/pipeline-config

Begin with resource_size: s and scale up if needed. This example sets the resource size to 's' with a comment to start small and monitor performance.

```yaml
name: my-pipeline
resource_size: s  # Start small, monitor performance
```

--------------------------------

### Clone and install dependencies

Source: https://docs.goldsky.com/compose/guides/build-a-polymarket-copy-trader

Commands to clone the example repository, navigate to the copy-trader directory, and install npm dependencies. Required before running `goldsky compose start` or `deploy`.

```bash
git clone https://github.com/goldsky-io/documentation-examples.git
cd documentation-examples/compose/copy-trader
npm install
```

--------------------------------

### Agent setup prompt for Goldsky

Source: https://docs.goldsky.com/introduction

Prompt for an AI assistant to install Goldsky agent skills, add the docs MCP server, install the CLI, and log in.

```text
Set up Goldsky for me: run `npx skills add goldsky-io/goldsky-agent` to install
the Goldsky agent skills, add the docs MCP server at https://docs.goldsky.com/mcp,
then install the Goldsky CLI (curl https://goldsky.com | sh) and log in with
`goldsky login`.
```

--------------------------------

### Clone the example repository

Source: https://docs.goldsky.com/compose/guides/build-a-vrf-system

Clones the Goldsky documentation examples repository and navigates to the VRF example directory.

```bash
git clone https://github.com/goldsky-io/documentation-examples.git
cd documentation-examples/compose/VRF
```

--------------------------------

### Install Goldsky AI skills via npx

Source: https://docs.goldsky.com/ai-skills

Recommended installation method. Installs to Claude Code, Cursor, and 30+ other AI agents.

```bash
npx skills add goldsky-io/goldsky-agent
```

--------------------------------

### Install Gill

Source: https://docs.goldsky.com/compose/guides/solana-transactions

Install the Gill library in your Compose project.

```bash
npm install gill
```

--------------------------------

### goldsky edge get <name> - example output

Source: https://docs.goldsky.com/reference/cli

Example output from the edge get command, showing endpoint details such as name, product, status, rate limit, allowed domains, and creation time.

```bash
Name:            grand-pulse
Product:         rpc
Status:          ACTIVE
Rate limit:      default
Allowed domains: unrestricted
Created at:      6/10/2026, 9:09:37 AM
```

--------------------------------

### Clone the example repository

Source: https://docs.goldsky.com/compose/guides/build-a-bitcoin-oracle

Commands to clone the documentation-examples repository and navigate to the bitcoin-oracle example directory.

```bash
git clone https://github.com/goldsky-io/documentation-examples.git
cd documentation-examples/compose/bitcoin-oracle
```

--------------------------------

### array_filter example: Filter Logs by Event Name

Source: https://docs.goldsky.com/turbo-pipelines/reference/sql-functions

Example using array_filter to get only 'Transfer' events from instruction logs.

```sql
sql: |
  SELECT
    signature,
    -- Get only 'Transfer' events from instruction logs
    array_filter(
      instruction_logs,
      'event_name',
      'Transfer'
    ) as transfer_events
  FROM solana_transactions
```

--------------------------------

### Clone the example repository

Source: https://docs.goldsky.com/compose/guides/build-a-nav-oracle

Commands to clone the documentation-examples repository and navigate to the compose/nav-oracle directory.

```bash
git clone https://github.com/goldsky-io/documentation-examples.git
cd documentation-examples/compose/nav-oracle
```

--------------------------------

### reverse_bytes32 example query

Source: https://docs.goldsky.com/turbo-pipelines/reference/sql-functions

Example SQL query using reverse_bytes32 on the transaction_hash column from the ethereum_transactions table to get the little-endian hash.

```sql
sql: |
  SELECT
    transaction_hash,
    reverse_bytes32(transaction_hash) as le_hash
  FROM ethereum_transactions
```

--------------------------------

### Clone the example repository

Source: https://docs.goldsky.com/compose/guides/build-a-corporate-actions-distributor

Commands to clone the documentation-examples repository and navigate to the corporate-actions compose example.

```bash
git clone https://github.com/goldsky-io/documentation-examples.git
cd documentation-examples/compose/corporate-actions
```

--------------------------------

### Source configuration with start_at: latest

Source: https://docs.goldsky.com/turbo-pipelines/guides/troubleshooting

Example source configuration showing start_at: latest, which only processes data created after deployment.

```yaml
sources:
  eth_transfers:
    type: dataset
    dataset_name: ethereum.erc20_transfers
    start_at: latest # Only new data from deploy time onward
```

--------------------------------

### Dataset Source Example (polygon_transfers)

Source: https://docs.goldsky.com/turbo-pipelines/pipeline-config

Example dataset source using the `matic.erc20_transfers` dataset version 1.2.0, starting at the latest block.

```yaml
sources:
  polygon_transfers:
    type: dataset
    dataset_name: matic.erc20_transfers
    version: 1.2.0
    start_at: latest
```

--------------------------------

### Enable Boost via CLI

Source: https://docs.goldsky.com/boost/quickstart

Run the interactive configuration flow to set up Boost. The wizard asks for a provider account, its management credential, and the networks to enable.

```bash
goldsky boost enable
```

--------------------------------

### Subgraph initialization complete!

Source: https://docs.goldsky.com/subgraphs/guides/create-a-no-code-subgraph

Final wizard summary after deployment, listing configuration files, build status, deploy status, and dashboard/query links.

```text
│
◇  Deploying 1 subgraphs
│
◇  nouns-demo-mainnet/1.0.0-demo+docs deployed
│
◇  Subgraph initialization summary
│
│  Configuration files:
│
│  • …/nouns-demo/1.0.0-demo+docs/abis/nouns.json
│  • …/nouns-demo/1.0.0-demo+docs/nouns-demo-mainnet-subgraph.json
│
│  Build:
│
│  ✔  BUILT mainnet
│
│  Deploy:
│
│  ✔  DEPLOYED nouns-demo-mainnet/1.0.0-demo+docs
│
┝━━━
│
◇  Deployed subgraph summary
│
│  nouns-demo-mainnet/1.0.0-demo+docs
│
│  • Dashboard: https://app.goldsky.com/…/dashboard/subgraphs/nouns-demo-mainnet/1.0.0-demo+docs
│  • Queries  : https://api.goldsky.com/api/public/…/subgraphs/nouns-demo-mainnet/1.0.0-demo+docs/gn
│
┝━━━
│
┕  Subgraph initialization complete!
```

--------------------------------

### Run goldsky command

Source: https://docs.goldsky.com/chains/0g

After logging in, run the goldsky command to get started.

```bash
goldsky
```

--------------------------------

### goldsky pipeline start - help output

Source: https://docs.goldsky.com/reference/cli

Shows the full help for the start command, including the --from-snapshot option which defaults to 'new' and the deprecated --use-latest-snapshot flag.

```bash
goldsky pipeline start <nameOrConfigPath>

Start a pipeline

Positionals:
  nameOrConfigPath  pipeline name or config path  [string] [required]

Options:
      --token                CLI Auth Token  [string] [default: ""]
      --color                Colorize output  [boolean] [default: true]
      --use-latest-snapshot  attempts to use the latest available snapshot.  [deprecated: Use '--from-snapshot'] [boolean]
      --from-snapshot        Snapshot that will be used to start the pipeline. Applicable values are: 'last', 'new', 'none' or a snapshot-id. 'last' uses latest available snapshot. 'new' creates a new snapshot to use. 'none': does not use any snapshot aka starts from scratch. Including the option without any argument will start an interactive mode to select from a list of available snapshots. Defaults to 'new'  [string]
  -h, --help                 Show help  [boolean]
```

--------------------------------

### _gs_generate_series SQL example

Source: https://docs.goldsky.com/turbo-pipelines/reference/sql-functions

Generates an array of integers from start to stop. The example expands one row per block in a range using CROSS JOIN UNNEST.

```sql
sql: |
  SELECT
    range_id || '_' || CAST(t.block_number AS STRING) as id,
    t.block_number
  FROM block_ranges
    CROSS JOIN UNNEST(
      _gs_generate_series(start_block, end_block)
    ) AS t (block_number)
```

--------------------------------

### Performance tip: Start with latest for development

Source: https://docs.goldsky.com/turbo-pipelines/sources/near

Use start_at: latest when building and testing pipelines to avoid processing large amounts of historical data.

```yaml
start_at: latest  # Only process new data

```

--------------------------------

### Basic GET Request

Source: https://docs.goldsky.com/compose/context/fetch

A minimal example of using `ctx.fetch` to perform a GET request and return the parsed JSON data. The response is automatically parsed as JSON with a text fallback.

```typescript
import { TaskContext } from "compose";

export async function main({ fetch }: TaskContext) {
  const data = await fetch("https://api.example.com/data");
  return data;
}
```

--------------------------------

### Task main Function Example

Source: https://docs.goldsky.com/compose/task-triggers

TypeScript example of a task's main function that accepts a payload with an optional name and returns a greeting. This is the function that gets called when the task is triggered.

```typescript
import { TaskContext } from "compose";

export async function main(_context: TaskContext, payload: { name?: string }) {
  const name = payload?.name || "World";

  return { success: true, greeting: `Hello, ${name}!` };
}
```

--------------------------------

### Proceed with subgraph build?

Source: https://docs.goldsky.com/subgraphs/guides/create-a-no-code-subgraph

Wizard prompt after writing configuration files, asking whether to proceed with the build stage.

```text
│
┕  Subgraph configuration complete!

┍  Initializing subgraph nouns-demo/1.0.0-demo+docs
│
◇  Writing subgraph files to '/Users/someone/my-subgraphs/nouns-demo/1.0.0-demo+docs': All subgraph configuration files written!
│
◆  Proceed with subgraph build?
│  ● Yes / ○ No
┕
```

--------------------------------

### Realtime Nodes (without system transactions) - viem

Source: https://docs.goldsky.com/edge-rpc/capabilities/hyperevm-system-transactions

viem example to get the latest block number from the realtime node.

```javascript
import { createPublicClient, http } from 'viem'

const client = createPublicClient({
  transport: http('https://edge.goldsky.com/standard/evm/999?key=demo&use-upstream=standard*')
})
const blockNumber = await client.getBlockNumber()
```

--------------------------------

### Realtime Nodes (without system transactions) - ethers.js

Source: https://docs.goldsky.com/edge-rpc/capabilities/hyperevm-system-transactions

ethers.js example to get the latest block number from the realtime node.

```javascript
import { JsonRpcProvider } from 'ethers'

const provider = new JsonRpcProvider(
  'https://edge.goldsky.com/standard/evm/999?key=demo&use-upstream=standard*'
)
const blockNumber = await provider.getBlockNumber()
```

--------------------------------

### Example database error: project size limit exceeded

Source: https://docs.goldsky.com/turbo-pipelines/guides/troubleshooting

Example log message indicating the database is out of storage (e.g., Neon's free tier caps at 512 MB). Upgrade the plan, clear old data, or add filtering to reduce write volume.

```text
Execution error: Failed to create table 'transfers': error returned from
database: could not extend file because project size limit (512 MB) has
been exceeded
```

--------------------------------

### Archive Nodes (with system transactions) - viem

Source: https://docs.goldsky.com/edge-rpc/capabilities/hyperevm-system-transactions

viem example to get block 1 with system transactions from the archive node.

```javascript
import { createPublicClient, http } from 'viem'

const client = createPublicClient({
  transport: http('https://edge.goldsky.com/standard/evm/999?key=demo&use-upstream=systx*')
})
const block = await client.getBlock({ blockNumber: 1n })
```

--------------------------------

### start

Source: https://docs.goldsky.com/compose/cli-reference

Starts the Compose app locally, preserving execution state from previous runs. Binds to port 4000 by default, walking up to 4009 if needed.

```APIDOC
## start

### Description
Starts your app locally. Preserves execution state from previous runs, useful for testing retry behavior and other production scenarios. By default the server binds port 4000; if that port is taken it walks up to 4009 and uses the first free port.

### Method
CLI command

### Command
`goldsky compose start`

### Options
- `-m, --manifest <path>` - Path to manifest file (default: `compose.yaml`)
- `--fork-chains` - Fork all chains referenced in contract interactions locally for testing
- `--impersonate <mappings>` - Impersonate wallet addresses on the local fork (e.g. `"wallet1=0xAddr1,wallet2=0xAddr2"`). Requires `--fork-chains`.
- `-p, --port <port>` - Bind exactly this port; fails fast if it is already in use (no walking)

### Example
```
goldsky compose start
```
```

--------------------------------

### Archive Nodes (with system transactions) - ethers.js

Source: https://docs.goldsky.com/edge-rpc/capabilities/hyperevm-system-transactions

ethers.js example to get block 1 with system transactions from the archive node.

```javascript
import { JsonRpcProvider } from 'ethers'

const provider = new JsonRpcProvider(
  'https://edge.goldsky.com/standard/evm/999?key=demo&use-upstream=systx*'
)
// Archival data with system transactions
const block = await provider.getBlock(1)
```

--------------------------------

### Proceed with subgraph deploy?

Source: https://docs.goldsky.com/subgraphs/guides/create-a-no-code-subgraph

Wizard prompt after the build stage, asking whether to proceed with the deploy stage.

```text
│
◇  Building subgraphs: 1 subgraph built!
│
◆  Proceed with subgraph deploy?
│  ● Yes / ○ No
┕
```

--------------------------------

### Deploy contracts with deploy.sh

Source: https://docs.goldsky.com/compose/guides/build-a-corporate-actions-distributor

Runs the deploy script to deploy all three contracts and pre-mint share-token balances to 25 demo addresses. Prints the three contract addresses and the ShareToken deploy block, which are needed for the next step.

```bash
PRIVATE_KEY=0x... ./scripts/deploy.sh
```

--------------------------------

### Solana (Blocks) Source Configuration

Source: https://docs.goldsky.com/turbo-pipelines/sources/overview

Example configuration for a Solana dataset source using blocks. Sets `start_block` to a specific slot number to start from that slot.

```yaml
sources:
  solana_blocks:
    type: dataset
    dataset_name: solana.blocks
    version: 1.0.0
    start_block: 312229952
```

--------------------------------

### Use start_at: latest for new pipelines

Source: https://docs.goldsky.com/turbo-pipelines/pipeline-config

Unless historical data is needed, start from the latest. This example sets start_at to 'latest' to only process new data.

```yaml
sources:
  my_source:
    start_at: latest  # Only process new data
```

--------------------------------

### Good example: dynamic table and transform using same source

Source: https://docs.goldsky.com/turbo-pipelines/transforms/dynamic-tables

Shows a valid configuration where both the dynamic table 'my_table' and the transform 'my_transform' reference source_1, ensuring data consistency.

```yaml
transforms:
  my_table:
    sql: SELECT value FROM source_1 # Uses source_1

  my_transform:
    sql: |
      SELECT * FROM source_1  # Also uses source_1
      WHERE dynamic_table_check('my_table', value)
```

--------------------------------

### Array Indexing for Nested Instructions

Source: https://docs.goldsky.com/turbo-pipelines/sources/solana

SQL examples for accessing individual instructions from the instructions array using 1-based indexing. Shows how to get the first instruction, its program_id, and data.

```sql
instructions[1]           -- First instruction
instructions[1].program_id -- First instruction's program
instructions[1].data      -- First instruction's data
```

--------------------------------

### Create a New Compose Project

Source: https://docs.goldsky.com/compose/quick-start

Run this command to create a new Compose project. It prompts for a project name and scaffolds a complete example app with a working Bitcoin oracle.

```bash
goldsky compose init
```

--------------------------------

### Column Transformations SQL Transform

Source: https://docs.goldsky.com/turbo-pipelines/transforms/sql

Transform column values using functions. The example lowercases addresses, casts value to DECIMAL and divides by 1e18 to get ETH, and selects block_timestamp.

```yaml
transforms:
  normalized_addresses:
    type: sql
    primary_key: id
    sql: |
      SELECT
        id,
        lower(from_address) as from_address,
        lower(to_address) as to_address,
        CAST(value AS DECIMAL) / 1e18 as value_eth,
        block_timestamp
      FROM transfers
```

--------------------------------

### Start Compose app with chain forking

Source: https://docs.goldsky.com/compose/debugging

Use the --fork-chains flag to start your app with in-memory forks of all chains you interact with, powered by TEVM. All wallets are automatically funded with test ETH, so you can test contract interactions without spending gas. Your task code stays the same as in production.

```bash
goldsky compose start --fork-chains
```

--------------------------------

### Configure start_at for Edge EVM dataset source

Source: https://docs.goldsky.com/turbo-pipelines/sources/edge-evm

Sets the starting block for an Edge EVM dataset source. start_at accepts 'latest', 'earliest', or a specific block number. The example uses a specific block number '18000000'.

```yaml
sources:
  eth_logs:
    type: dataset
    dataset_name: edge.evm.logs
    version: 1.0.0
    start_at: "18000000" # latest | earliest | a specific block number
    chain_id: 1
```

--------------------------------

### Example: Decode System Program Instructions

Source: https://docs.goldsky.com/turbo-pipelines/reference/sql-functions

Shows a SQL transform that decodes System Program instructions using `_gs_solana_decode_system_program_instruction`.

```yaml
transforms:
  system_decoded:
    type: sql
    primary_key: id
    sql: |
      SELECT
        id,
        block_slot,
        signature,
        _gs_solana_decode_system_program_instruction(data, accounts) as decoded
      FROM solana_ix
      WHERE program_id = '11111111111111111111111111111111'
```

--------------------------------

### Example: tuning a 12-partition topic

Source: https://docs.goldsky.com/turbo-pipelines/sinks/kafka

YAML configuration for a Kafka sink with parallelism set to 4, batch_size 50000, and batch_flush_interval 500ms. Use this as a starting point for tuning a 12-partition topic. Adjust parallelism to a value that evenly divides the number of partitions.

```yaml
sinks:
  kafka_output:
    type: kafka
    from: high_volume_events
    topic: events.stream
    data_format: avro
    secret_name: MY_KAFKA_SECRET
    topic_partitions: 12
    primary_key: id
    parallelism: 4          # Divides evenly into 12 partitions
    batch_size: 50000
    batch_flush_interval: 500ms
```

--------------------------------

### Install OpenZeppelin contracts with Foundry

Source: https://docs.goldsky.com/compose/guides/build-a-compliance-oracle

Installs the OpenZeppelin contracts library, required for the escrow contract's IERC20 import.

```bash
forge install OpenZeppelin/openzeppelin-contracts --no-commit
```
Tokens:
Raw