github.com/eng-floe/floecat @main

Floecat is a catalog-of-catalogs for modern data lakehouses. It federates metadata harvested from Delta Lake and Iceberg sources, stores a canonical view in append-only blob storage, and exposes the resulting hierarchy over gRPC for discovery, authorization, and query planning.

The repository is purposely modular. Each top-level directory corresponds to an operational component (service runtime, connector packages, storage backends, CLI, etc.). Detailed, component-specific documentation lives under docs/.

Quick Start (Docker + LocalStack)

Baseline requirements for quickstart: Docker Engine + Docker Compose v2.

Run Floecat using published GHCR images (currently :main) and a LocalStack backend:

FLOECAT_SERVICE_IMAGE=ghcr.io/eng-floe/floecat-service:main \
FLOECAT_ICEBERG_REST_IMAGE=ghcr.io/eng-floe/floecat-iceberg-rest:main \
FLOECAT_CLI_IMAGE=ghcr.io/eng-floe/floecat-cli:main \
FLOECAT_PULL_POLICY=always \
FLOECAT_ENV_FILE=./env.localstack \
COMPOSE_PROFILES=localstack \
docker compose -f docker/docker-compose.yml up -d

This starts:

• ghcr.io/eng-floe/floecat-service:main
• ghcr.io/eng-floe/floecat-iceberg-rest:main
• ghcr.io/eng-floe/floecat-cli:main
• LocalStack (COMPOSE_PROFILES=localstack)

Stop the stack:

FLOECAT_SERVICE_IMAGE=ghcr.io/eng-floe/floecat-service:main \
FLOECAT_ICEBERG_REST_IMAGE=ghcr.io/eng-floe/floecat-iceberg-rest:main \
FLOECAT_CLI_IMAGE=ghcr.io/eng-floe/floecat-cli:main \
FLOECAT_PULL_POLICY=always \
FLOECAT_ENV_FILE=./env.localstack \
COMPOSE_PROFILES=localstack \
docker compose -f docker/docker-compose.yml down --remove-orphans

Use the CLI after the stack is up:

# attach to the running cli container
FLOECAT_SERVICE_IMAGE=ghcr.io/eng-floe/floecat-service:main \
FLOECAT_ICEBERG_REST_IMAGE=ghcr.io/eng-floe/floecat-iceberg-rest:main \
FLOECAT_CLI_IMAGE=ghcr.io/eng-floe/floecat-cli:main \
FLOECAT_PULL_POLICY=always \
FLOECAT_ENV_FILE=./env.localstack \
COMPOSE_PROFILES=localstack \
docker compose -f docker/docker-compose.yml attach cli

# or start a fresh one-off CLI session
FLOECAT_SERVICE_IMAGE=ghcr.io/eng-floe/floecat-service:main \
FLOECAT_ICEBERG_REST_IMAGE=ghcr.io/eng-floe/floecat-iceberg-rest:main \
FLOECAT_CLI_IMAGE=ghcr.io/eng-floe/floecat-cli:main \
FLOECAT_PULL_POLICY=always \
FLOECAT_ENV_FILE=./env.localstack \
COMPOSE_PROFILES=localstack \
docker compose -f docker/docker-compose.yml run --rm --use-aliases cli

In the CLI, set the demo account t-0001 before issuing catalog queries:

floecat> account t-0001
account set: 5eaa9cd5-7d08-3750-9457-cfe800b0b9d2
floecat> catalogs
CATALOG_ID                                CREATED_AT                DISPLAY_NAME              DESCRIPTION
ab5bcc89-fcad-3574-935f-b5249eacf700      2026-02-25T13:29:00.947Z  examples                  Examples catalog
floecat>

Detach without stopping the CLI container:

Ctrl+P, Ctrl+Q

Optional Make shortcuts:

make quickstart-up
make quickstart-down

# override owner/tag
make quickstart-up QUICKSTART_OWNER=<owner> QUICKSTART_TAG=<tag>

DuckDB (host) with LocalStack

From the repository root, start DuckDB with the LocalStack bootstrap script:

duckdb -init tools/duckdb-localstack-init.sql

The init script installs/loads required DuckDB extensions, configures S3 to use LocalStack, and attaches the Floecat Iceberg REST catalog as iceberg_floecat.

Example sanity checks:

SHOW DATABASES;
SELECT COUNT(*) FROM iceberg_floecat.delta.call_center;
INSERT INTO iceberg_floecat.iceberg.barf VALUES (1);

If you launch DuckDB outside the repo root, pass the absolute init script path.

Build, Test, and Run

Contributor requirements: Java 25+, Maven, Make.

make build
make test
make run

Seed data is enabled by default (floecat.seed.enabled=true); the service starts with a demo account, catalogs, namespaces, tables, and snapshots.

Documentation

• Architecture, system flow, and modules: docs/architecture.md
• Storage model and key layout: docs/storage-layout.md
• API access patterns: docs/api-access.md
• External authentication and session headers: docs/external-authentication.md
• System scan overview (gRPC + Flight): docs/system-scans.md
• gRPC system scan protocol: docs/system-scans-grpc.md
• Arrow Flight protocol and integration: docs/arrow-flight.md
• CLI commands: docs/cli-reference.md
• Operations, testing, and observability: docs/operations.md

• Docker compose modes and overrides: docs/docker.md

• Example catalog extension (file-based, zero-code): docs/extension-example.md

All additional component docs live under docs/.

Contributing

Contribution workflow, review policy, and contributor expectations are documented in:

• CONTRIBUTING.md

All participants are expected to follow:

• CODE_OF_CONDUCT.md
• SECURITY.md for vulnerability reporting