DATABASES / dbt Labs
dbt Documentation Maintenance
Audits descriptions in a dbt manifest and drafts missing model or column documentation in the project’s existing style.
“Audits dbt documentation coverage and drafts missing model/column descriptions in the project's own house style, one folder at a time, for human review. …”
01 / THE REASONING
Why this made the selection.
- Uses dbt’s resolved manifest rather than guessing YAML layout, and clearly limits column coverage to columns declared in that manifest.
- Requires tracing grain and column meaning through SQL, refs and sources before writing descriptions, then leaves a scoped diff for human review.
02 / THE REVIEW RECORD
What we actually inspected.
Source review has boundaries.
A clear record is more useful than a “safe” badge.
Material inspected
- skills/dbt/skills/maintaining-dbt-documentation/SKILL.md
- skills/dbt/skills/maintaining-dbt-documentation/audit_coverage.py
- skills/dbt/skills/using-dbt-for-analytics-engineering/references/writing-documentation.md
- README.md (installation, compatibility and licence sections)
- LICENSE
- Skill directory (SKILL.md, audit_coverage.py), the companion writing-documentation.md and LICENSE compared with the previous review revision a8607fc0: byte-identical at 77923f56 (git blob SHAs). README.md adds a dbt-packages installation section.
Our findings
- The entire audit_coverage.py helper was inspected. It reads manifest JSON locally, groups model nodes by original_file_path and counts stripped nonempty descriptions.
- The workflow refreshes the manifest, re-parses after changes, preserves existing correct descriptions and lists meanings it could not establish.
- The helper itself performs no network calls or file writes; the surrounding dbt workflow can generate artifacts and optional warehouse inspection needs separate credentials.
- Its frontmatter limits tools to Bash(dbt *), Bash(python3 *), Bash(git *), Read, Write, Edit, Glob and Grep, and sets user-invocable: false.
Not established by this review
- The helper was not executed and no dbt project or warehouse was accessed.
- No generated description was checked with a domain owner.
- Not executed: it was not installed or run.
The review applies to the material and revision named here. A newer upstream release can change its behavior.
03 / PUT IT TO WORK
Add a skill to your workflow.
DOCUMENTED COMMAND
npx skills add dbt-labs/dbt-agent-skills/skills/dbt/skills/maintaining-dbt-documentation Run in your terminal to choose your AI tools, project or global scope, and installation method. This short command follows the current upstream version; the review covers the revision shown above.
Optional: install the reviewed revision
Keeps the original source and installer version where recorded, while leaving your AI tools and installation scope for you to choose.
npx skills add https://github.com/dbt-labs/dbt-agent-skills/tree/77923f561356ab83ebd467c79c6805868346b33d/skills/dbt/skills/maintaining-dbt-documentation - Install the complete directory, including audit_coverage.py.
- Keep the working directory at the dbt project root and use the helper’s full installed path or its explicit manifest option.
- The publisher’s running-dbt-commands and analytics-engineering skills provide optional companion guidance; install them if using those handoffs.
- Run the folder audit for each intended directory; review new files and out-of-folder edits as well as tracked diffs.
- The README at this revision also documents installing the skills as dbt packages through packages.yml with revision: main and an ai_provider flag in dbt_project.yml; it advises pinning revision to a tag or commit SHA for reproducible installs.
- Choose your AI tools and project or global scope in the installer. The source revision shown here is the one reviewed; the installer itself is not executed or version-pinned by this review.
Before you start
- Node.js/npm and git for the documented Skills installer
- A configured dbt project and its correct dbt executable
- Python for the bundled audit helper; warehouse access only for optional physical-column inspection
04 / FOLLOW THE EVIDENCE
The source trail.
Our notes are separate from the original resource.
Check upstream before adopting a new version.
https://github.com/dbt-labs/dbt-agent-skills/blob/77923f561356ab83ebd467c79c6805868346b33d/skills/dbt/skills/maintaining-dbt-documentation/SKILL.md
Supports: summary, upstreamDescription, whySelected, bestFor, limitations, review, access
https://github.com/dbt-labs/dbt-agent-skills/blob/77923f561356ab83ebd467c79c6805868346b33d/skills/dbt/skills/maintaining-dbt-documentation/audit_coverage.py
Supports: review, limitations, install, access
https://github.com/dbt-labs/dbt-agent-skills/blob/77923f561356ab83ebd467c79c6805868346b33d/skills/dbt/skills/using-dbt-for-analytics-engineering/references/writing-documentation.md
Supports: whySelected, review, limitations
https://github.com/dbt-labs/dbt-agent-skills/blob/77923f561356ab83ebd467c79c6805868346b33d/README.md
Supports: install, compatibility, access, license
https://github.com/dbt-labs/dbt-agent-skills/blob/77923f561356ab83ebd467c79c6805868346b33d/LICENSE
Supports: license, access.cost
https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/README.md
Supports: install
https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/src/git.ts
Supports: install
https://github.com/vercel-labs/skills/blob/13e4063a1cf913f5606d57d42ab83a86f5001e04/src/source-parser.ts
Supports: install
https://raw.githubusercontent.com/dbt-labs/dbt-agent-skills/77923f561356ab83ebd467c79c6805868346b33d/skills/dbt/skills/maintaining-dbt-documentation/SKILL.md
Supports: upstreamDescription, review.notTested, review.reviewedAt