Tools Reference

All tools require an authenticated workspace context and return JSON text. Tools are read-only and cannot mutate workspace state. List-style tools cap limit at 100 rows.

get_workspace_overview

Returns workspace name, target brand, active prompt/location/engine counts, tracked competitors, and the latest completed measurement window.

Input: none.

list_prompts

Lists tracked prompts with text, intent, tier, topic, prompt group, and active state.

Inputs:

• active boolean, optional.
• search string, optional case-insensitive prompt text search.
• limit number from 1 to 100, optional.

list_measurement_windows

Lists tracking snapshots newest first within the plan history window.

Inputs:

• status string, optional. One of planned, running, completed, failed.
• limit number from 1 to 100, optional.

get_visibility_overview

Returns headline visibility metrics per brand and engine for a measurement window. Defaults to the latest completed window.

Inputs:

• measurement_window_id string, optional. Defaults to latest completed measurement window.
• brand_id string, optional.

get_prompt_visibility

Returns per-engine and per-location visibility metrics for one prompt.

Inputs:

• query_id string, required. Use list_prompts to find it.
• measurement_window_id string, optional. Defaults to latest completed measurement window.

list_citations

Returns cited domains grouped by domain with citation counts and sample URLs.

Inputs:

• measurement_window_id string, optional. Defaults to latest completed measurement window.
• domain string, optional canonical domain filter.
• limit number from 1 to 100, optional.

list_recommendations

Returns optimization recommendations by priority.

Inputs:

• status string, optional.
• category string, optional.
• limit number from 1 to 100, optional.

list_competitor_visibility

Compares the target brand and tracked competitors for a measurement window, sorted by visibility score.

Inputs:

• measurement_window_id string, optional. Defaults to latest completed measurement window.

list_crawler_events

Returns daily AI crawler visits captured from your tracked sites (GPTBot, ClaudeBot, PerplexityBot, Bingbot, and more). Each row is one day's hit count for a crawler on a path, split by HTTP status class when the install reports it, with a last_seen_at timestamp. Requires crawler tracking to be installed on at least one site. For path-level filtering or AI-referral rows, use the REST resource: AI Crawler Events.

Inputs:

• crawler_id string, optional. Filter to one crawler, e.g. gptbot.
• site_key_id string, optional. Filter to one tracked site.
• limit number from 1 to 100, optional.

list_claims

Returns Brand FactCheck claims extracted from AI answers about your tracked brands, each verified against the workspace's approved facts. Every row carries its verdict, error class, confidence, theme, and review state. Mirrors the REST resource: Claims.

Inputs:

• run_id string, optional. Filter by run id.
• brand_id string, optional. Filter by brand id.
• engine_id string, optional, e.g. openai_web.
• verdict string, optional. One of accurate, inaccurate, unverifiable.
• theme_id string, optional. Filter by claim theme id.
• review_state string, optional. One of auto, needs_review, confirmed, corrected, suppressed.
• limit number from 1 to 100, optional.

For example, filter to verdict=inaccurate with engine_id=openai_web to pull just the inaccurate claims one engine made in the latest windows.

get_accuracy_overview

Returns the per-engine Accuracy Score for the latest completed measurement window, each engine's delta against the previous completed window, an overall score across engines, and fact knowledge-base coverage. Use it for a one-call health check without paging through accuracy aggregates.

Input: none.

The response includes latest_completed_window, previous_completed_window, overall_accuracy_score, a by_engine array (each with accuracy_score, previous_accuracy_score, delta, and the verdict counts), and a knowledge_base object with active, proposed, and stale fact counts.