# XDRCheck > A test kit for software that reads the Stellar ledger: a corpus of rare but valid XDR shapes tagged by CAP, a generator, a replay harness with adapters and service rigs, an independent judge, an SDK conformance runner, a linter for Go indexers and an indexer-hardening guide. - [Introduction](https://docs.xdrcheck.dev/index.md): Test software that reads the Stellar ledger against rare but valid transaction shapes, each tagged with the CAP that introduced it and emitted by a real stellar-core. - [Quickstart](https://docs.xdrcheck.dev/quickstart.md): Read the corpus offline from a fresh clone, then fetch once, build, and run the demo, the judge and find. - [For SDK maintainers](https://docs.xdrcheck.dev/start/for-sdk-maintainers.md): The short path for an SDK or XDR-library team: copy a recipe, pin a corpus release, commit the list of failures you expect, and every corpus entry goes through your decoder on each pull request. - [For indexer authors and operators](https://docs.xdrcheck.dev/start/for-indexers.md): The short path for an indexer team: build the kit, replay the corpus through the adapter template, read the report, then point the adapter at your own indexer and run it in CI. - [How to read a corpus entry](https://docs.xdrcheck.dev/start/read-an-entry.md): One real entry walked field by field: CAP and protocol tags, the synthetic flag, both passphrases, the expected decoded values and the re-encode check. - [Architecture overview](https://docs.xdrcheck.dev/architecture/overview.md): The problem XDRCheck solves in Stellar terms, the design decisions and why they matter to an SDK maintainer or an indexer operator, and how the parts fit together. - [The corpus: bytes and expected values](https://docs.xdrcheck.dev/architecture/corpus.md): Where every corpus byte comes from, how expected values are established without any reader under test, the round-trip and golden checks, and one real entry walked through. - [Mainnet scan and the mainnet tier](https://docs.xdrcheck.dev/architecture/mainnet-scan.md): How every pubnet ledger from 2 to 65,505,999 was read and checked, the shape-token grammar that describes real traffic, the 6,158-transaction tier that holds every transaction-level token, and what the scan does not claim. - [Replay: corpus ledgers through your indexer](https://docs.xdrcheck.dev/architecture/replay.md): How ledgers reach an unmodified indexer through the SEP-54 lake or getLedgers, what the adapter reports, how each entry's verdict is computed from the rows the indexer stored, and how wallet-backend, stellar-etl and Horizon are run. - [Conformance: the corpus through your SDK](https://docs.xdrcheck.dev/architecture/conformance.md): How xdrcheck conform drives one SDK version per column, what one cell checks at each level, how known divergences are declared and when a run fails, and how a defect found in one SDK becomes a check on the others. - [Linter and hardening guide](https://docs.xdrcheck.dev/architecture/linter-and-guide.md): What xdrlint detects in Go ingestion code built on go-stellar-sdk, how its results on wallet-backend and stellar-etl are pinned, and how the indexer-hardening guide is kept equal to the corpus. - [CI: tiers, offline runs and what each job proves](https://docs.xdrcheck.dev/architecture/ci.md): How every check runs offline after one fetch, how versions are pinned, the pull-request and full CI tiers job by job, the manual nightly and watch workflows, measured times, release artifacts, and how to wire the kit into your own CI. - [Stellar integration](https://docs.xdrcheck.dev/architecture/stellar-integration.md): Every Stellar component XDRCheck builds on or tests: the version pinned in pins.toml, what the kit consumes from it, what it gives back, and where the integration lives. - [Protocol upgrades and maintenance](https://docs.xdrcheck.dev/architecture/protocol-upgrades.md): Protocol 30 coverage, how the kit fails loudly on a newer protocol, how it sees an upgrade coming, the upgrade procedure step by step, and what an SDK or an indexer gets at each upgrade. - [Using the corpus in an SDK's CI](https://docs.xdrcheck.dev/guides/sdk-ci.md): Put every corpus entry through your SDK's decoder in its own CI: loaders, recipes and expected-failure lists. - [Protocol upgrades: what XDRCheck gives an SDK](https://docs.xdrcheck.dev/guides/protocol-upgrade-sdk.md): Test a new protocol's shapes before the SDK release. - [Minimal adapter: the template](https://docs.xdrcheck.dev/guides/go-indexer.md): Replay the corpus through a Go indexer with the adapter template: run it, copy it, fill in the marked places. - [A minimal replay adapter in TypeScript](https://docs.xdrcheck.dev/guides/any-language-indexer.md): The adapter template in TypeScript, for an indexer in any language. - [xdrlint](https://docs.xdrcheck.dev/guides/lint-go.md): Install xdrlint, run it, read a diagnostic, suppress with a reason and run it in CI. - [CI recipes](https://docs.xdrcheck.dev/guides/ci.md): Run XDRCheck in your CI: the GitHub Action for indexers, SDKs and Go linting, expected-failure lists, replaying pinned real ledgers through an SDK branch, and running offline with bounded memory. - [Add corpus entries for a CAP](https://docs.xdrcheck.dev/guides/add-corpus-entries.md): Register an entry ID, produce the bytes from a local stellar-core, a golden, network history or a recipe, write what a correct reader must derive, and prove it with verify. - [Maintenance and protocol upgrades](https://docs.xdrcheck.dev/guides/maintain-kit.md): Versioning, releases and channels, the protocol-upgrade procedure, who maintains the kit and for how long. - [Indexer-hardening guide](https://docs.xdrcheck.dev/hardening/overview.md): What the indexer-hardening guide and its checklist cover, where they live, and how they are kept consistent with the corpus and the linter. - [Outline and review state](https://docs.xdrcheck.dev/hardening/outline.md): The guide's sections by topic, its checklist by mistake and how each item is checked, and how many of its claims a stellar-core or platform engineer has reviewed. - [Verification report: indexer-hardening guide](https://docs.xdrcheck.dev/hardening/verification.md): How every statement of the indexer-hardening guide was checked: evidence classes, pinned versions, the claims that matter most, and what is still open. - [Results](https://docs.xdrcheck.dev/results/overview.md): What the runs show for SDK maintainers and indexer operators: the mainnet scan, the corpus, incidents reproduced, the SDK matrix and the findings, each with its page and the command that reproduces it. - [Mainnet scan: integrity check and shape census](https://docs.xdrcheck.dev/results/mainnet-scan.md): Every pubnet ledger from 2 to 65,505,999: what was checked and the result, the shape census in exact terms, prevalence per CAP and per entry, the mainnet tier's token coverage, how to reproduce it and what is not claimed. - [SDK conformance matrix](https://docs.xdrcheck.dev/results/sdk-matrix.md): Every pinned Go, JavaScript, Python and Rust SDK release against the corpus: agrees, diverges and not run per release, what a divergence means, and what the current releases still get wrong. - [Replay results](https://docs.xdrcheck.dev/results/replay.md): wallet-backend, stellar-etl and the example indexer against the corpus: what each class means for your data, results per version and input, three defects before and after their fixes, and what the current versions get wrong. - [Known incidents](https://docs.xdrcheck.dev/results/incidents.md): Public incidents in Stellar SDKs, XDR libraries and indexers, the corpus entries that hold each shape, and whether the kit reproduces it before and after the fix, runs it, checks it or holds it as input. - [Findings](https://docs.xdrcheck.dev/results/findings.md): Every public defect record by project and severity, with the corpus entries that reproduce it and its upstream state; and each defect class checked across SDK families. - [CAPs, patterns and ledgers](https://docs.xdrcheck.dev/caps/overview.md): For every CAP and coverage group: the corpus entries, the pattern each holds, its protocol, where its bytes come from, and how often rare union arms occur on pubnet. - [Command-line reference](https://docs.xdrcheck.dev/reference/cli.md): Every command of the kit's tools on one page: the conventions they share, what each command does and its full --help. - [make targets](https://docs.xdrcheck.dev/reference/make-targets.md): Every make target: what it runs, whether it needs the network, and who it is for. - [Manifest and index format](https://docs.xdrcheck.dev/reference/manifest.md): The corpus manifest and index: entries, variants, sources, expected values, validity, conformance levels and entry IDs. - [Replay adapter protocol — `xcc.adapter/1`](https://docs.xdrcheck.dev/reference/adapter-protocol.md): xcc.adapter/1: the JSON messages between the replay harness and an adapter that drives your indexer. - [Replay harness — `xdrcheck serve` and `xdrcheck replay`](https://docs.xdrcheck.dev/reference/replay-harness.md): How the harness serves ledgers, judges what an indexer stored, classifies each entry, and runs the service rigs; the baseline and stellar-etl adapters and the Horizon rig. - [`xdrcheck replay --target`: any target, any version, any ledger](https://docs.xdrcheck.dev/reference/replay-target.md): xdrcheck replay --target: an SDK or pipeline at any version against chosen ledgers, transactions, sets, CAPs or issues; the dataset, find, the run directory and repro bundles. - [Conformance driver protocol and rules](https://docs.xdrcheck.dev/reference/driver-protocol.md): xcc.driver/1, the messages between the conformance runner and a driver that runs one SDK version; the runner's outcomes, evidence and judging rules; xcheck and mutation. - [Judge (xdrcheck-judge)](https://docs.xdrcheck.dev/reference/judge.md): The judge: ground truth from a ledger's own entry changes, the row schema, scope files and verdicts. - [xdrlint rules](https://docs.xdrcheck.dev/reference/lint-rules.md): Every xdrlint rule on one page: what it reports, why it matters, examples, limits and how to suppress it; how to run the linter, its output formats and configuration. - [Measured run times](https://docs.xdrcheck.dev/reference/ci-timings.md): Measured run times and disk use of the make targets, CI jobs and fetch parts. - [Infrastructure, user tracking and decentralisation](https://docs.xdrcheck.dev/reference/statements.md): What the kit reaches over the network, what it records, what it depends on, and how to check each. - [Threat model and monitoring plan](https://docs.xdrcheck.dev/reference/threat-model.md): Threat model of the kit and its monitoring plan. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.