Occupations
Read-only occupation reference: occupations, job zones, skills, abilities, work styles, interest types, related occupations, job titles and software. Use get-occupation-profile for one complete occupation instead of combining separate endpoints. Data is static per release and cached. Every response carries a mandatory attribution field that must be shown on every publicly visible display.
Try it with your AI assistant
Once WebsitePublisher is connected to your AI assistant, paste one of these into the chat. Replace the parts in [brackets] with your own details.
@WebsitePublisher build a career page where visitors search an occupation and see its key skills, with the O*NET attribution shown@WebsitePublisher match the job titles from this CV to occupations and list the closest matches: [job titles]With WebsitePublisher, build a career page where visitors search an occupation and see its key skills, with the O*NET attribution shownWith WebsitePublisher, match the job titles from this CV to occupations and list the closest matches: [job titles]With WebsitePublisher, build a career page where visitors search an occupation and see its key skills, with the O*NET attribution shownWith WebsitePublisher, match the job titles from this CV to occupations and list the closest matches: [job titles]Endpoints (6)
Get one complete occupation: title, description, job zone with explanation, all descriptor values and related occupations. Pass onetsoc_code in O*NET-SOC format, for example "15-1252.00". The answer contains a coverage object that says per domain whether there was data — not every occupation has every descriptor, and a missing domain is a fact, not an error.
| Field | Type | Required | Description |
|---|---|---|---|
onetsoc_code |
string | ✓ Yes | O*NET-SOC code in the format NN-NNNN.NN, for example "15-1252.00" (Software Developers). |
domains |
array | No | Which descriptor domains to include. Allowed: essential_skills, transferable_skills, abilities, knowledge, work_activities, work_styles, career_interest_types. Default: essential_skills, transferable_skills, abilities, work_styles, career_interest_types. |
include_related |
string | No | Set to "false" to leave out related occupations. They are included by default. |
Match free-text terms against the occupation taxonomy and get candidate elements back. Meant for skill extraction: send the terms that come from a CV, get grounded taxonomy elements back. This is a lexical match on whole words in element name and description; terms of 1-2 characters only match exactly. Each term gives exactly one result — a term that cannot be matched comes back with the reason in skipped. It is a list of candidates, not a decision. Which candidate counts is decided in the scoring layer, not here.
| Field | Type | Required | Description |
|---|---|---|---|
terms |
array | ✓ Yes | List of terms to match. At most 50 per call, each at most 100 characters. |
domains |
array | No | Limit the match to these domains. Allowed: essential_skills, transferable_skills, abilities, knowledge, work_activities, work_styles, career_interest_types. Default: essential_skills and transferable_skills. |
limit_per_term |
string | No | Maximum number of candidates per term, 1 to 25. Default 10. |
Match job titles (for example from a CV) and get candidate occupations back, one per occupation: onetsoc_code, occupation title and the title that matched best. The search covers the job titles, O*NET's short titles ("IT Manager", "RN") and the occupation titles themselves; matched_field says which source matched. Lexical word overlap: an identical title scores 1.0, and "Senior Software Developer" also finds "Software Developer". Terms of 1-2 characters only match exactly. Each term gives exactly one result — a term that cannot be matched comes back with the reason in skipped. It is a list of candidates, not a decision.
| Field | Type | Required | Description |
|---|---|---|---|
terms |
array | ✓ Yes | List of job titles. At most 50 per call, each at most 100 characters. |
limit_per_term |
string | No | Maximum number of candidates per term, 1 to 25. Default 10. |
Match software names (for example from a CV) against the software list and get the example back with its software category (element_id and element_name from the content model), the number of occupations that mention it, hot_technology and the number of occupations it is in demand for. Whole words: "Java" does not find "JavaScript". Names and terms of 1-2 characters, such as "C" or "R", only match exactly. Each term gives exactly one result — a term that cannot be matched comes back with the reason in skipped. It is a list of candidates, not a decision.
| Field | Type | Required | Description |
|---|---|---|---|
terms |
array | ✓ Yes | List of software names. At most 50 per call, each at most 100 characters. |
limit_per_term |
string | No | Maximum number of candidates per term, 1 to 25. Default 10. |
Browse or search the occupation list. Returns code, title and description. Use search to filter on title.
| Field | Type | Required | Description |
|---|---|---|---|
search |
string | No | Search term on the occupation title. At least 2 characters. |
page |
string | No | Page number, from 1. Default 1. |
per_page |
string | No | Number per page, 1 to 100. Default 25. |
Return the loaded release: label, load date, source URL, sha256 of the source archive and which files were deliberately not loaded. Use this to anchor calculation traces — a stored score should be able to point to the release data it was calculated against.
No input parameters required.
MCP Tool Names
When using this integration through an AI assistant (Claude, ChatGPT, Cursor, etc.), the endpoints are available as MCP tools:
| Endpoint | MCP Tool Name |
|---|---|
| get-occupation-profile | occupations_get_occupation_profile |
| match-elements | occupations_match_elements |
| match-titles | occupations_match_titles |
| match-software | occupations_match_software |
| list-occupations | occupations_list_occupations |
| get-release | occupations_get_release |
Website