--- name: ingest-sdk description: Build custom network ingestion pipelines on Stellar. Covers the Go Ingest SDK, history archives, ledger metadata, ApplyLedgerMetadata, Galexie exports, and real-time or historical data flows. user-invocable: true argument-hint: "[ingest task]" --- # Stellar Ingestion Pipelines > Build custom network ingestion pipelines on Stellar. Use the Go Ingest SDK to read ledger metadata, derive application models, and stream data to downstream systems. ## When to use this skill - Building a pipeline that consumes live Stellar ledger metadata - Exporting historical or continuous ledger data with Galexie - Translating XDR metadata into a domain model - Sending derived events to a datastore or message bus - Choosing between a custom indexer, Galexie, and RPC/Hubble workflows ## Related skills - Historical query and RPC/Horizon workflows → `../data-indexers/SKILL.md` - Reading live network state or events → `../rpc/SKILL.md` - Writing consumers that react to contract events → `../soroban/SKILL.md` --- ## What the Ingest SDK is The Ingest SDK is a set of Go packages under `github.com/stellar/go-stellar-sdk` for acquiring and parsing Stellar network data. It converts XDR-encoded streams from Stellar Core into typed data structures that are easier to process in application code. ## Quick reference | Package | Purpose | |---------|---------| | `amount` | Converts operation prices and amounts into strings | | `historyarchive` / `datastore` / `storage` | Accesses history archives and datastores | | `ingest` | Parses ledger metadata and applies it to consumer callbacks | | `network` | Pre-configured network constants for Testnet/Mainnet | | `xdr` | Go bindings for the Stellar network data model | ## Typical pipeline An end-to-end ingestion flow usually has two parts: 1. **Export pipeline** — exports Stellar ledger metadata to cloud storage. 2. **Consumer pipeline** — reads exported files and turns them into an app-specific model. ### Consumer pipeline roles - **Inbound adapter** — reads `LedgerCloseMeta` files from storage and feeds them into `ApplyLedgerMetadata`. - **Transformer** — parses ledger metadata with `xdr`, filters the operations you care about, and maps them to your model. - **Outbound adapter** — publishes the derived model to a datastore, queue, or message bus. ## Galexie Use **Galexie** to export Stellar ledger metadata to cloud storage. Common modes: ```bash --start --end --start ``` Galexie is the first step in Stellar's Composable Data Platform for building historical data lakes. ## Source of truth - [Build Custom Network Ingestion Pipeline](https://developers.stellar.org/docs/build/apps/ingest-sdk) - [Ingest SDK](https://developers.stellar.org/docs/data/indexers/build-your-own/ingest-sdk) - [Galexie](https://developers.stellar.org/docs/data/indexers/build-your-own/galexie) - [BufferedStorageBackend](https://developers.stellar.org/docs/data/indexers/build-your-own/ingest-sdk/developer_guide/ledgerbackends/bufferedstoragebackend) ## See also - `/data-indexers/SKILL.md` — query historical data and indexers - `/rpc/SKILL.md` — Stellar RPC for real-time state --- *raw.githubusercontent.com/ggoldani/stellarskills/main/ingest-sdk — MIT License*