Understand the boundaries before trying to read everything

Arrange twelve repositories by role and choose one to read today.

Suggested learning time: 30 minutes.

1. 1Reader question
2. 2Repository role
3. 3Product and ownership boundaries
4. 4One feature to inspect

Original learning map: arrows show the reading or decision sequence, not a measured execution trace.

Status of this chapter

This textbook was authored from reading sources. No actual repository builds, tests, debugging, API calls, or issue/PR posts were performed. Practical work below is an exercise to undertake; reading is not an execution result.

About this course

This workbook covers reading and verifying Codex and Databricks-related open source, then preparing a contribution that fits each project’s policy.

Eight chapters trace one feature with evidence instead of requiring you to understand a huge repository all at once. Learning and public contribution are separate outcomes.

For learners with basic Git knowledge and one programming language. Suggested total learning time: 5.75 hours. The first catalog curates twelve repositories without claiming completeness.

Prerequisites:

• Git commits, branches, and diffs.
• Functions, types, and exceptions.
• Terminal basics.

Learning objectives

• Arrange twelve repositories by role and choose one to read today.

Do not aim for a complete collection

These twelve entries are the first catalog, not a list of every related open-source project. “Related” includes official implementations, official clients, dependent or adjacent projects, and community extensions. First identify four things: users, inputs, outputs, and external services. Do not read a huge repository in file order.

A map of the landscape

For Codex, you can read public code on the terminal side. For Databricks, the CLI, management SDKs, and SQL Connector are separate entry points. Spark is a computation engine; Delta Lake handles tables and storage; MLflow covers tracking, evaluation, and other aspects of AI development; Unity Catalog OSS is a catalog project. Do not call all of them official Databricks repositories.

Your first choice

Start with the Python SDK if you know Python, the CLI if you want to trace processing from configuration, or Codex if you are interested in AI tools. One repository is enough for the first week. Understanding all of Spark is not a prerequisite for a small SDK reproduction experiment.

Initial twelve-repository catalog

These twelve repositories are an initial selection. The official, adjacent, and community categories can be extended, but the initial twelve do not include community repositories. This is not an exhaustive list, popularity ranking, or claim that every implementation was reviewed. Cards were reviewed on 2026-10-03; checking a README and public structure is distinct from reading all source code.

Category	How to establish the relationship
Official	Implementations, clients, or SDKs published by OpenAI or Databricks.
Adjacent OSS	Separately owned projects related by dependencies, connections, or the surrounding ecosystem.
Community	Third-party extensions; verify evidence of the relationship when adding an entry.

Submitting a PR is not the Codex learning outcome. The current official contribution guide does not accept external code contributions or PRs. Proceed toward reproduction, cause analysis, and an issue draft, and recheck the policy immediately before any public post.

openai/codex

ID: codex · Reviewed: 2026-10-03

Canonical repository: openai/codex

• Classification: Official OpenAI
• Languages: Rust / TypeScript
• Learning fit: First priority: reading and cause analysis
• Contribution policy and caveats: External code contributions and PRs are currently not accepted. Prepare reproduction and cause-analysis drafts.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Reviewed scope: codex-rs/app-server/README.md — The protocol documentation was reviewed. This does not mean the entire application is publicly available.

Source ID: codex-app.

Read the verified external Codex DeepWiki entry. Verification here concerns link availability, not correctness of its explanations.

databricks/cli

ID: cli · Reviewed: 2026-10-03

Canonical repository: databricks/cli

• Classification: Official Databricks
• Languages: Go
• Learning fit: First priority: Bundle processing
• Contribution policy and caveats: Development procedures were reviewed. Recheck the acceptance policy before a public contribution.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Reviewed scope: bundle/bundle.go — The Bundle/Mutator descriptions and SDK import in the file were reviewed.

Source ID: cli-bundle.

databricks/databricks-sdk-py

ID: sdk-py · Reviewed: 2026-10-03

Canonical repository: databricks/databricks-sdk-py

• Classification: Official Databricks
• Languages: Python
• Learning fit: A first practical candidate: the HTTP boundary
• Contribution policy and caveats: A public mirror. PRs receive public review and are applied internally. Signature and DCO requirements apply.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Reviewed scope: databricks/sdk/core.py — The constructor's configuration passing to _BaseClient was reviewed.

Source ID: sdk-core.

databricks/databricks-sdk-go

ID: sdk-go · Reviewed: 2026-10-03

Canonical repository: databricks/databricks-sdk-go

• Classification: Official Databricks
• Languages: Go
• Learning fit: Understand layers beneath the CLI
• Contribution policy and caveats: Reread the contribution guide before posting.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

databricks/databricks-sdk-java

ID: sdk-java · Reviewed: 2026-10-03

Canonical repository: databricks/databricks-sdk-java

• Classification: Official Databricks
• Languages: Java
• Learning fit: Further study on the JVM
• Contribution policy and caveats: Reread the contribution guide before posting.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

databricks/databricks-sql-python

ID: sql-python · Reviewed: 2026-10-03

Canonical repository: databricks/databricks-sql-python

• Classification: Official Databricks
• Languages: Python
• Learning fit: The SQL client boundary
• Contribution policy and caveats: Learn how its purpose differs from a management SDK.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

databricks/terraform-provider-databricks

ID: terraform · Reviewed: 2026-10-03

Canonical repository: databricks/terraform-provider-databricks

• Classification: Official Databricks
• Languages: Go
• Learning fit: Infrastructure as code and state management
• Contribution policy and caveats: Operations on real resources require separate checks of costs and permissions.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

mlflow/mlflow

ID: mlflow · Reviewed: 2026-10-03

Canonical repository: mlflow/mlflow

• Classification: Adjacent OSS: independent project
• Languages: Python / TypeScript
• Learning fit: A contribution candidate for AI evaluation and tracking
• Contribution policy and caveats: The current guide recommends consulting through an issue about the implementation approach.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

apache/spark

ID: spark · Reviewed: 2026-10-03

Canonical repository: apache/spark

• Classification: Adjacent OSS: Apache project
• Languages: Scala / Java / Python
• Learning fit: The execution engine and Spark Connect
• Contribution policy and caveats: Do not label this as a repository owned by Databricks.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

delta-io/delta

ID: delta · Reviewed: 2026-10-03

Canonical repository: delta-io/delta

• Classification: Adjacent OSS: Delta Lake project
• Languages: Scala / Java / Python
• Learning fit: Table formats and transactions
• Contribution policy and caveats: Study its boundary with the computation engine.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

delta-io/delta-rs

ID: delta-rs · Reviewed: 2026-10-03

Canonical repository: delta-io/delta-rs

• Classification: Adjacent OSS: Delta Lake project
• Languages: Rust / Python
• Learning fit: Language boundaries
• Contribution policy and caveats: The contribution-guide link was verified, but retrieval of its body was not completed.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

unitycatalog/unitycatalog

ID: unitycatalog · Reviewed: 2026-10-03

Canonical repository: unitycatalog/unitycatalog

• Classification: Adjacent OSS: LF AI & Data sandbox project
• Languages: Java and other languages
• Learning fit: Catalogs
• Contribution policy and caveats: Do not equate it with commercial Databricks Unity Catalog. Check AI_POLICY too.
• Next single action: Choose one README usage example and explain its input, output, and external boundaries in four lines.
• Issue availability: Not investigated. No guarantee that an issue is open, unassigned, or suitable for beginners.
• Version status: Reviewed on a rolling branch; pin the SHA when practicing.

Contribution/development guide: [source URL retained in private provenance]

Pinned commit SHA: null (unverified). Reviewing a rolling branch is not verification of a pinned version.

Implementation-reading scope, codeAreas: []. Do not fill unread paths or detailed architecture with guesses.

When extending the catalog

Retain the canonical URL, publisher, evidence for the relationship, license, archive status, contribution policy, and verification date. Values absent from the initial cards remain unverified. Link to official material rather than redistributing third-party code or DeepWiki output.

Continue to the next chapter to pin a version

Thirty-five primary references and review scope

The verification date records reading a source, rather than publication, update, or execution. Links can change later; recheck material at the same commit as your practice version.

• openai/codex official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• openai/codex contribution and development guide — 2026-10-03: External code contributions and PRs are currently not accepted. Prepare reproduction and cause-analysis drafts.
• databricks/cli official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• databricks/cli contribution and development guide — 2026-10-03: Development procedures were reviewed. Recheck the acceptance policy before a public contribution.
• databricks/databricks-sdk-py official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• databricks/databricks-sdk-py contribution and development guide — 2026-10-03: A public mirror: PRs receive public review and are applied internally. Signature and DCO requirements apply.
• databricks/databricks-sdk-go official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• databricks/databricks-sdk-go contribution and development guide — 2026-10-03: Reread the contribution guide before posting.
• databricks/databricks-sdk-java official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• databricks/databricks-sdk-java contribution and development guide — 2026-10-03: Reread the contribution guide before posting.
• databricks/databricks-sql-python official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• databricks/databricks-sql-python contribution and development guide — 2026-10-03: Learn how its purpose differs from a management SDK.
• databricks/terraform-provider-databricks official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• databricks/terraform-provider-databricks contribution and development guide — 2026-10-03: Real-resource operations require separate checks of costs and permissions.
• mlflow/mlflow official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• mlflow/mlflow contribution and development guide — 2026-10-03: The current guide recommends discussing an implementation approach through an issue.
• apache/spark official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• apache/spark contribution and development guide — 2026-10-03: Do not label this as a repository owned by Databricks.
• delta-io/delta official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• delta-io/delta contribution and development guide — 2026-10-03: Study its boundary with the computation engine.
• delta-io/delta-rs official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• delta-io/delta-rs contribution and development guide — 2026-10-03: The contribution-guide link was verified, but retrieval of its body was incomplete.
• unitycatalog/unitycatalog official repository — 2026-10-03: The README and public structure were reviewed; this is not a close reading of all source code.
• unitycatalog/unitycatalog contribution and development guide — 2026-10-03: Do not equate it with commercial Databricks Unity Catalog. Check AI_POLICY too.
• Bundle implementation — 2026-10-03: The Bundle/Mutator descriptions and SDK import were reviewed.
• ApiClient implementation — 2026-10-03: The constructor’s configuration passing to _BaseClient was reviewed.
• Codex app-server documentation — 2026-10-03: The protocol documentation was reviewed. This does not mean the whole application is publicly available.
• Codex build guide — 2026-10-03: Recheck instructions at the same commit when practicing.
• Official DeepWiki MCP — 2026-10-03: An unauthenticated MCP for public repositories; /mcp is recommended.
• Official DeepWiki documentation — 2026-10-03: A specification permitting iframe embedding was not confirmed.
• Cognition Platform Terms — 2026-10-03: Updated 2026-06-30. Applicability to public MCP and permission to redistribute remain unresolved.
• Cognition Privacy Policy — 2026-10-03: Check the scope of data transmission before integration.
• Databricks Connect — 2026-10-03: A Databricks-oriented extension based on Spark Connect.
• Apache Spark Connect — 2026-10-03: Client-server separation.
• Databricks and open source — 2026-10-03: An entry point for checking relationships between company products and open source.

Practical exercise: Practice: understand the boundaries first

1. Choose one repository.
2. Describe users, inputs, outputs, and external services in four lines.
3. Open the README and contribution guide.
4. Choose two areas you will not read this time.

Deliverable: One role card.

Understanding checks

Answer in your own words before reading the answers. The answers and explanations below are for self-review, not automatic scores or execution results.

Are Databricks Connect and Spark Connect the same?

Answer: Databricks Connect is a Databricks-oriented extension built on Spark Connect. Separate the open-source implementation from the commercial product.

Your learning progress

Recording mode: manual.

• Not started
• Read
• Tried
• Can explain

Viewing a page alone does not mean you completed the practical exercise.

Material used in this chapter

openai/codex official repository, databricks/cli official repository, databricks/databricks-sdk-py official repository, apache/spark official repository, Databricks Connect, Databricks and open source

Return to repository entry points · Offline experiment worksheet

Current Learn edition

The chapter set was restored to the common Learn archive on 2026-10-04. Original source reviews retain their 2026-10-03 date. The former shared experiment screen is not available in this edition: use the linked offline worksheet and keep every planned run marked not-run until observations exist. No experiment, ranking, or performance measurement was created by this restoration.