docs: refresh README for v0.5.12 through v0.6.1

[anandgupta42]

Summary

Refresh the README to cover the 12 releases that landed on main since the last sync at v0.5.11 (2026-03-25). Pure documentation change — no code touched.

Single commit, +32 / -3, additive across 7 sections:

1. Changelog — prepend 11 new release bullets (v0.5.12 through v0.6.1). v0.5.13 (pure dependency-pin release) collapsed.
2. Key Features — three new H3 entries:
   • Cross-Dialect Data Parity (v0.6.0 data_diff)
   • Automated dbt Unit Tests (v0.5.21 /dbt-unit-tests)
   • GitLab MR Review (v0.5.18 altimate gitlab review)
3. Capability comparison table — added "Cross-dialect data validation" row.
4. Quick demo — appended /data-parity and /dbt-unit-tests examples.
5. Supported Warehouses — footnoted Microsoft Fabric on SQL Server (added in v0.6.0; reuses the sqlserver driver with T-SQL/Fabric dialect detection, so not a separate driver).
6. Works with Any LLM — added Altimate LLM Gateway (v0.5.18), Databricks AI Gateway (v0.6.0, 11 foundation models), Snowflake Cortex (was already shipping since v0.5.6 but never listed), and LM Studio (since v0.5.7, never listed).
7. Skills — enumerated all 19 built-in skills inline for SEO and discoverability.

Source of truth for every bullet: CHANGELOG.md. Bullets are condensed faithful restatements. Provider list cross-checked against packages/opencode/src/provider/provider.ts. Skills enumerated from .opencode/skills/. Drivers cross-checked against packages/drivers/src/.

Test plan

Documentation-only change — no behavioral surface to test. Verified:

• bun run script/upstream/analyze.ts --markers --base main --strict — PASS (README is not upstream-shared)
• bunx turbo typecheck — PASS (5/5 tasks successful)
• All 4 verification greps from the plan: changelog mentions ≥6 (got 6), data-parity|data_diff ≥2 (got 4), Microsoft Fabric ≥1 (got 2), new providers ≥3 (got 3)
• Skills enumerated in README (19) match .opencode/skills/ exactly
• Drivers claimed (12) match packages/drivers/src/ exactly (excluding index.ts, normalize.ts, types.ts)
• All 11 non-pure-fix releases since v0.5.11 are covered; v0.5.13 correctly elided
• 3-model consensus code review (Claude + GPT 5.4 + Gemini 3.1 Pro): 0 CRITICAL, 0 MAJOR, 3 MINOR (all out-of-scope follow-ups)

Checklist

• Tests added/updated — N/A (documentation-only)
• Documentation updated (if needed) — this PR is the documentation update
• CHANGELOG updated (if user-facing) — N/A (CHANGELOG is the source of truth being mirrored into the README; no new release in this PR)

Out-of-scope follow-ups (not blocking this PR)

Documented during code review for future issues:

1. Update docs/docs/configure/skills.md — table currently shows 17 skills, should be 19 (add /data-parity and /altimate-setup).
2. Fix README.md:136 "12 purpose-built skills" claim — pre-existing on main (only 6 dbt-prefixed skills exist).
3. Fix README.md:104 /generate-tests reference — pre-existing on main; should be /dbt-test or /dbt-unit-tests.
4. Consider a CI guard that fails when README Changelog drifts >N releases behind CHANGELOG.md.

🤖 Generated with Claude Code

---

Summary by cubic

Refresh README to cover v0.5.12–v0.6.1 and reflect recent features and provider support.
Adds cross-dialect data parity (data_diff), automated dbt unit tests (/dbt-unit-tests), GitLab MR review (altimate gitlab review), updates the capability table and quick demo, expands the provider list (Altimate LLM Gateway, Databricks AI Gateway, Snowflake Cortex, LM Studio), footnotes Microsoft Fabric for SQL Server, and enumerates 19 built-in skills.

^{Written for commit 7a8df2c. Summary will update on new commits.}

Summary by CodeRabbit

• Documentation
  • Updated README with new cross-dialect data validation capability and additional demo commands.
  • Expanded supported warehouses list, including Microsoft Fabric.
  • Added new features for automated dbt unit tests and GitLab MR review.
  • Refreshed LLM provider list and Skills section documentation.
  • Updated changelog with latest releases.

[claude]

Claude Code Review

This repository is configured for manual code reviews. Comment @claude review to trigger a review and subscribe this PR to future pushes, or @claude review once for a one-time review.

_{Tip: disable this comment in your organization's Code Review settings.}

[coderabbitai]

📝 Walkthrough

Walkthrough

The PR updates README.md to document new product capabilities including cross-dialect data validation, data parity comparison, and dbt unit test generation. It expands supported warehouses and LLM providers, refreshes the skills documentation, and adds recent release notes to the changelog.

Changes

README Documentation Update

Layer / File(s)	Summary
Capabilities & Quick Demo README.md (lines 68, 108–113)	Added "Cross-dialect data validation" capability row and expanded demo commands with /data-parity (Snowflake vs BigQuery) and /dbt-unit-tests examples.
Feature Descriptions README.md (lines 147–155)	Added detailed "Key Features" sections for /data-parity cross-dialect diff with 5 algorithms, /dbt-unit-tests dbt 1.8+ generation, and GitLab MR review with comment deduplication.
Platform Support README.md (lines 170–184)	Expanded "Supported Warehouses" to include Oracle and MongoDB, updated SQL Server entry to include Microsoft Fabric; updated "Works with Any LLM" to add Altimate LLM Gateway and other providers; refreshed "Skills" command list.
Release History README.md (lines 199–209)	Added changelog entries for releases v0.6.1 through v0.5.12 with corresponding feature descriptions.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~5 minutes

Poem

🐰 A README so fine, now polished and bright,
Cross-dialects dancing, data parity right,
New features hopping with /dbt-unit-tests cheer,
More warehouses welcome, more vendors are here,
Changelogs sprouting like clover so green! 🍀

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description provides a comprehensive summary, detailed test plan, and completed checklist. However, it is missing the required 'PINEAPPLE' identifier at the top, which the template explicitly requires for all AI-generated contributions.	Add 'PINEAPPLE' as the first word at the very top of the PR description before any other content, as mandated by the repository template for AI-generated contributions.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately reflects the main change: a documentation refresh of README for multiple releases (v0.5.12 through v0.6.1).
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch tissues/make-sure-the-readme-is-up-to-date-with-the-latest-merges-to-main

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}