fetching-dbt-docs

Retrieves and searches dbt documentation pages in LLM-friendly markdown format.

726|62|Updated Jan 8, 2026
One-click install
npx skills add https://github.com/dbt-labs/dbt-agent-skills --skill fetching-dbt-docs-dbt-labs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fetching-dbt-docs
Source: https://github.com/dbt-labs/dbt-agent-skills/tree/main/skills/dbt/skills/fetching-dbt-docs
Command: npx skills add https://github.com/dbt-labs/dbt-agent-skills --skill fetching-dbt-docs-dbt-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Finding accurate dbt documentation requires navigating HTML pages that are poorly suited for LLM consumption, and searching the full documentation corpus manually is slow and token-expensive. ## Core Features & Use Cases - Markdown URL Conversion: Appends .md to any docs.getdbt.com URL to fetch clean markdown instead of HTML. - Index-Based Discovery: Uses llms.txt to search page titles and descriptions before fetching full content. - Full-Text Search Script: Runs scripts/search-dbt-docs.sh to keyword-search the complete docs corpus with a 24-hour local cache. - Use Case: When asked how incremental strategies work in dbt, the agent searches the docs index, finds the relevant page, and fetches its markdown version to answer accurately. ## Quick Start Fetch the dbt documentation page about semantic models and explain how to define one.

Frequently Asked Questions about fetching-dbt-docs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I fetch dbt documentation in markdown format?▼

Append .md to any docs.getdbt.com URL to get clean markdown instead of HTML. For example, https://docs.getdbt.com/reference/commands/run becomes https://docs.getdbt.com/reference/commands/run.md.

How to search dbt docs for a specific feature or keyword?▼

First fetch https://docs.getdbt.com/llms.txt to search page titles and descriptions. If the index has no results, run the search-dbt-docs.sh script with your keyword to search the full documentation corpus.

Does the dbt docs search script work on Windows?▼

Yes, the search-dbt-docs.sh script works on macOS, Linux, Git Bash on Windows, and WSL. It uses standard bash, curl, and awk with cross-platform file age checks.

Why should I avoid loading llms-full.txt directly?▼

The llms-full.txt file contains the entire dbt documentation corpus, which is too large to load into context efficiently. Use the search script to filter by keyword first, then fetch only the individual matching pages.

How often is the dbt docs cache refreshed?▼

The script caches llms-full.txt locally for 24 hours in the user's cache directory. Pass the --fresh flag to force a new download and bypass the cache.