Okf Bigquery

BigQuery OKF Connector

This skill provides a Go-based CLI tool to convert Google Cloud BigQuery dataset schemas and table/field descriptions into Open Knowledge Format (OKF) bundles, and synchronize comments/descriptions back to BigQuery tables and fields.

When to Use

Use this skill when you need to:

1. Extract table structures, schemas, field descriptions, and dataset descriptions from a Google Cloud BigQuery dataset into a portable OKF bundle.
2. Synchronize business descriptions or documentation changes written in an OKF bundle back into native BigQuery table and schema field descriptions.

Setup

The connector requires Go 1.24+ and utilizes the official Google Cloud BigQuery Client:

# Install the published binary (Go 1.24+) — no clone needed:
go install github.com/xSAVIKx/okf-skills/skills/okf-bigquery@v0.1.0

# …or build from a clone of the repository:
cd skills/okf-bigquery
go build -o okf-bigquery .

How to Use

Ensure you have set your GCP credentials. Usually, this is done by setting the GOOGLE_APPLICATION_CREDENTIALS environment variable:

export GOOGLE_APPLICATION_CREDENTIALS="/path/to/credentials.json"

1. Produce an OKF Bundle

Extract dataset schema metadata and descriptions:

./okf-bigquery produce --project  --dataset  --out  [--tables ] [--sample ] [--profile] [--relationships] [--stats]

2. Ingest / Synchronize an OKF Bundle

Synchronize table and field descriptions from the OKF bundle back to BigQuery:

./okf-bigquery ingest --project  --dataset  --bundle  [--sync]

Parameters:

• --project (required): Google Cloud Project ID.
• --dataset (required): Target BigQuery dataset ID.
• --bundle (required for ingest): Path to the OKF bundle.
• --out (required for produce): Path to output OKF bundle.
• --tables (optional): Filter to extract only specific tables.
• --sample (optional): Embed up to N sample rows per table as a ## Sample section in each table doc (default 0 = none).
• --profile (optional): Compute per-column statistics (non-null, null, distinct, min, max) and embed a ## Data Profile section.
• --relationships (optional): Extract informational foreign-key constraints into a ## Relationships section.
• --stats (optional): Add a ## Stats section with the table row count (from table metadata) and, when a timestamp-like column is detected, a freshness window (min/max). Constraints (## Constraints) and view definitions (## View Definition, views only) are emitted by default. BigQuery has no secondary indexes, so no Indexes section is produced.
• --sync (optional for ingest): If provided, calls the BigQuery API to update table and schema descriptions.

3. Inspect the Schema (self-description)

Print a machine-readable JSON description of this skill's commands and flags (used by okf-mcp to expose the skill as an MCP tool):

./okf-bigquery schema

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: xSAVIKx
• Source: xSAVIKx/okf-skills
• License: Apache-2.0
• Homepage: https://xsavikx.github.io/okf-skills/

Install and usage instructions live in the source repository linked above.