REST API

Inspect the API key and the active organization, together with the abilities the key currently holds and the projects it may act in.

An organization is the top-level billing entity and tenant boundary in Pineprompt. Every project, member, monitor, prompt, brand, and source belongs to exactly one organization.

A project represents a single brand or site you would like to monitor. Each project owns its own prompts, brands, monitors, and sources.

A monitor is a configured daily run. Each monitor binds a set of prompts to one or more AI engines in a chosen country and language. While a monitor is active, Pineprompt runs it once per day on your behalf.

A prompt is a single question that the monitor asks each engine on every run. Each prompt belongs to exactly one monitor.

Brands are the entities matched against engine answers: your own brand and any competitors you would like to track. Brand matches are what drive share of voice, sentiment, and citation tracking. Brands themselves are created by Pineprompt; only their classification is editable.

Tags are user-defined labels you may attach to prompts to classify them, for example `branded`, `non-branded`, or `review`. Every new project is seeded with these three defaults, and you may filter visibility metrics on any tag when querying analytics. Tag names are unique within a project and case-insensitive.

Segments group brands into named cohorts, such as `Direct Competitors` or `Enterprise Competitors`. A single brand may belong to multiple segments, and you may use segments to filter the dashboard. Every new project is seeded with a `Direct Competitors` segment, and a project must always have at least one segment. Only brands classified as `own` or `competitor` may be attached to a segment.

A source is a URL that an engine cited inside an answer. Pineprompt tracks sources over time so that you may see which pages drive your visibility. Sources are read-only.

A response is a single rendered answer from an AI engine to a single prompt on a single run. Each response captures the raw text, the engine that produced it, and the country and language the monitor was run in. Responses are read-only.

A brand mention is a single occurrence of a tracked brand inside an AI response, recorded together with its position in the answer and its sentiment score. Brand mentions are read-only.

A citation is a single URL that an AI engine linked to inside a response. Pineprompt aggregates citations into sources, which lets you see which pages drive your visibility over time. Citations are read-only.

A source URL is a single page that an AI engine cited inside an answer. Each source URL belongs to a source, the domain that issued it. Pineprompt tracks each source URL over time so that you may see which individual pages drive your visibility. Source URLs are read-only, and you will encounter one as part of a citation or a source analytics row rather than from a dedicated endpoint.

Aggregated metrics computed across the project's analyzed responses. Every analytics endpoint accepts a `start_date` and `end_date` together with a shared set of filter axes (models, countries, monitors, segments, and tags), and returns either a time series bucketed by `interval` (day, week, month, or year) or an aggregate list. When the window includes today, the result is cached until the next UTC midnight; otherwise it is cached for 24 hours.