| name | recommend-study-filters |
| description | Given a description of your target participant population, fetches all available Prolific filters and recommends the best matching combination with filter IDs and values ready to use. |
| version | 0.1.1 |
Recommend Study Filters
When asked to recommend filters for a Prolific study, follow these steps.
Step 1: Fetch all available filters
Mandatory: run the following Prolific CLI command to understand the filters CLI command.
prolific filters --help
Mandatory: run the following Prolific CLI command to retrieve the full filter catalogue and renders it straight to the terminal for analysis
prolific filters -n
Each filter in the output includes:
- Filter ID — the identifier used when configuring a study (e.g.
current-country-of-residence, age)
- Type — either
select (choose from a list of options) or range (set lower/upper bounds)
- Choices — for
select filters, a map of choice keys to human-readable labels
- Min / Max — for
range filters, the allowed bounds
Step 2: Analyse the researcher's requirements
Read the researcher's description carefully and identify every distinct requirement. Group them by category, for example:
- Geographic — country of residence, nationality
- Demographic — age, gender, ethnicity
- Educational — highest qualification, student status
- Employment — employment status, occupation, industry
- Behavioural / experiential — prior Prolific participation, device ownership, health conditions, parental status, language fluency
Flag any requirements that are ambiguous or may not map to a single filter.
Step 3: Match requirements to filters
For each identified requirement, find the best matching filter from the Step 1 output:
- Compare the researcher's language to filter titles, descriptions, and questions
- For
select type filters: identify the specific choice key(s) that match (e.g. for a "UK residents" requirement, find the key for "United Kingdom" in the country filter's choices)
- For
range type filters: determine appropriate lower and upper bounds based on the researcher's stated criteria (e.g. age 25–45 → lower: 25, upper: 45)
- Note the exact
filter_id for every recommended filter
If multiple filters could satisfy the same requirement (e.g. nationality vs. country of residence), explain the difference and recommend the more appropriate one.
Step 4: Present recommendations
For each recommended filter, provide:
- Filter title and filter_id
- Selected values (for
select type) or range (for range type), ready to copy into a study configuration
- A plain-language explanation of why this filter matches the researcher's requirement
Format the selected values in the JSON structure used by the Prolific API so they can be used directly:
- Select filter:
{ "filter_id": "...", "selected_values": ["key1", "key2"] }
- Range filter:
{ "filter_id": "...", "selected_range": { "lower": X, "upper": Y } }
If any stated requirement could not be matched to an available filter, say so explicitly and suggest the closest alternative or a workaround.