Combining Operators for Precision
Stacking filters into one sharp query
You'll be able to
- Stack type:, app:, and updated: filters in a single query
- Read a layered query left to right and predict its result set
Read first
Day 7 Lab
- Start with a broad keyword that returns far too much.
- Add
type:to keep one kind of document. - Add
app:to keep one connected system. - Add
updated:to keep one recency window. - Read the full query back and confirm each filter earned its place.
Working Example: Stacking Filters
Each operator removes a different slice of noise. Stack them and the result set narrows to exactly what you meant. Verified against search_filters.
budget → thousands of mixed results
type:document budget → only documents
type:document app:drive budget → only Drive documents
type:document app:drive updated:past_week budget → only Drive docs changed this week
Read it left to right: type cuts the kind, app cuts the source, updated cuts the age.Check for understanding
What does each operator in a stacked query remove from the result set?
Check yourself
1. What does `roadmap updated:past_week` return?
2. How do you isolate results from Q1 2026 with a precise date window?
3. What is the difference between `in:` and `collection:`?
4. When is `my:history pricing` the right move?
5. After Glean Assistant gives a cited answer, what is the governed next step?
In 10 seconds
“One filter narrows; three carve. Stack type:document app:drive updated:past_week budget and each operator strips away another slice until only the exact set remains. This is how precise search is built.”
Consuming Built-in Glean Skills