Sanity GROQ Query Skill

Write, debug, and optimize GROQ queries for Sanity CMS with the v2026-02-19 parser.

When to use this skill

• Writing new GROQ queries for Sanity data fetching
• Debugging GROQ parse errors
• Optimizing query performance
• Adding new document types or fields

Sanity project details

• Project ID: 4bg9jxr4
• Dataset: production
• API Version: 2026-02-19
• API Host: apicdn.sanity.io

Critical: GROQ v2026-02-19 parser quirks

The 2026-02-19 parser is stricter than older versions:

Field ordering matters

// WRONG — parser sees galleryItems as part of the outer object
*[_type == "project" && slug.current == $slug][0]{
  title,
  slug,
  galleryItems{
    images[]{ url, alt }
  },
  technologies // parser error here
}

// CORRECT — nested projections must come AFTER all top-level fields
*[_type == "project" && slug.current == $slug][0]{
  title,
  slug,
  technologies,
  tier,
  showcaseDetail,
  galleryItems{
    images[]{ url, alt }
  }
}

Use quoted field names for aliases

// WRONG — unquoted aliases cause parse errors
*[_type == "project"]{
  title,
  projectTitle: title // error
}

// CORRECT — use quoted names for aliases
*[_type == "project"]{
  title,
  "projectTitle": title
}

Array projections need careful nesting

// WRONG
*[_type == "project"]{
  title,
  galleryItems[]{
    images[]{ url, alt, caption }
  }
}

// CORRECT
*[_type == "project"]{
  title,
  galleryItems[] {
    images[] { url, alt, caption }
  }
}

Query patterns

Fetch all projects (summary)

*[_type == "project"] | order(orderRank asc) {
  title,
  slug,
  description,
  tier,
  technologies,
  orderRank
}

Fetch single project by slug

*[_type == "project" && slug.current == $slug][0] {
  title,
  slug,
  description,
  tier,
  projectUrl,
  repoUrl,
  orderRank,
  featured,
  technologies,
  features,
  challenges,
  learnings,
  metrics,
  showcaseDetail {
    challenge,
    solution,
    result,
    architecture,
    highlights
  },
  galleryItems[] {
    images[] { url, alt, caption, type }
  },
  _updatedAt
}

Fetch hero section

*[_type == "heroSection"][0] {
  name,
  title,
  roleRotator,
  contactEmail
}

Fetch blog posts

*[_type == "blogPost"] | order(publishedAt desc) {
  title,
  slug,
  publishedAt,
  excerpt,
  tags,
  readingTime,
  featured
}

Fetch certifications

*[_type == "certification"] | order(issuedDate desc) {
  name,
  issuer,
  issuedDate,
  expirationDate,
  credentialId,
  credentialUrl,
  logo {
    asset->{url}
  }
}

Fetch experience

*[_type == "experience"] | order(startDate desc) {
  company,
  position,
  location,
  startDate,
  endDate,
  current,
  description,
  highlights,
  technologies
}

Fetch skills

*[_type == "skill"] | order(category asc, name asc) {
  name,
  category,
  level,
  yearsOfExperience
}

Common errors

"Parse error: expected }"

• Usually a missing comma between fields
• Check nested projections for missing closing braces

"Parse error: unexpected token"

• Unquoted alias names
• Missing commas between fields
• Gallery/nested projection before top-level fields

"Cannot query"

• Wrong dataset name
• Missing * at the start of query
• Wrong _type name

Debugging workflow

1. Test query in Sanity Vision (Sanity Studio)
2. Use console.log to see raw query string
3. Check for comma placement
4. Verify field names match schema
5. Ensure nested projections come last

Checklist

• Query works in Sanity Vision
• No parse errors with v2026-02-19
• Nested projections come after top-level fields
• Aliases use quoted names
• Slug input is sanitized (no GROQ injection)