Discovery Abilities

Discovery abilities are read-only — they explain your site without changing anything, which makes them the perfect starting point. Most of the everyday value of Accelerate AI lives here: performance summaries, content rankings, experiment results, and traffic breakdowns, all answerable in plain language.

accelerate/get-performance-summary

DiscoveryRead

Returns a site-level performance summary for a date preset or custom date range, including traffic, conversion, and engagement metrics.

Ask the AI

How did my site perform over the last 7 days?

ParameterTypeRequiredDescription
date_range_presetstringNoOne of 1h, 4h, 12h, 24h, 7d, 30d, or 90d.

Example Request

json
{
  "name": "accelerate/get-performance-summary",
  "arguments": { "date_range_preset": "7d" }
}

REST Equivalent

bash
curl -X POST /wp-json/wp-abilities/v1/abilities/accelerate%2Fget-performance-summary/run \
  -H "Content-Type: application/json" \
  -d '{"input":{"date_range_preset":"7d"}}'

accelerate/list-active-experiments

DiscoveryRead

Lists active A/B tests with status, traffic allocation, variants, and current sample counts.

Ask the AI

Which experiments are currently running?

accelerate/list-experiments

DiscoveryRead

Lists historical and active experiments with filtering by status, type, date range, post, and annotations. Supports pagination.

Ask the AI

What have we tested on the homepage so far?

accelerate/get-audience-segments

DiscoveryRead

Lists configured audience segments so an agent can understand available targeting groups.

accelerate/get-audience-fields

DiscoveryRead

Returns fields available for audience creation, including geography, UTM, device, behavior, and custom attributes.

accelerate/get-top-content

DiscoveryRead

Ranks content by traffic, engagement, conversion, or another supported metric over the selected period.

Ask the AI

Which posts drove the most conversions this month?

accelerate/get-content-diff

DiscoveryRead

Compares two content items or two time windows to show what changed in traffic and engagement.

accelerate/get-traffic-breakdown

DiscoveryRead

Breaks traffic down by dimensions such as device, referrer, region, or campaign.

accelerate/get-post-performance

DiscoveryRead

Returns performance metrics for a single post, page, or block using an ID, URL, or resolved content match.

accelerate/get-taxonomy-performance

DiscoveryRead

Groups analytics by category, tag, or custom taxonomy to identify strong verticals.

accelerate/get-source-breakdown

DiscoveryRead

Shows traffic, engagement, and conversion grouped by source, medium, and campaign.

accelerate/get-utm-performance

DiscoveryRead

Analyzes UTM campaign performance, including conversion contribution rather than clicks alone.

accelerate/get-landing-pages

DiscoveryRead

Ranks entry pages by first-session traffic, conversion, and downstream engagement.

accelerate/get-attribution-comparison

DiscoveryRead

Compares first-touch and last-touch attribution for channels, campaigns, or content.

accelerate/get-concurrent-visitors

DiscoveryRead

Returns a live concurrent visitor estimate for real-time monitoring and spike investigation.

accelerate/get-author-performance

DiscoveryRead

Compares contributors or desks by traffic, engagement, conversion, and publishing velocity.

accelerate/get-author-content

DiscoveryRead

Lists an author’s content catalog with per-post performance and velocity.

accelerate/get-engagement-metrics

DiscoveryRead

Returns bounce rate, scroll depth, time on page, engaged time, recirculation, and other engagement metrics for content.

accelerate/get-experiment-results

DiscoveryRead

Returns Bayesian A/B test results: conversion rates, each variant’s probability to win, credible intervals, and a recommendation on whether to ship.

Ask the AI

Did the homepage test actually work?

accelerate/get-variants

DiscoveryRead

Lists variants for an experiment or personalization rule, including labels, current state, and the raw block markup for each variant so an agent can read the content it is about to test.

accelerate/get-site-context

DiscoveryRead

Returns site-level context an agent needs before suggesting experiments or changes.

accelerate/search-content

DiscoveryRead

Finds posts, pages, or blocks by title, URL, or text so other abilities can target the right object.

accelerate/get-content

DiscoveryRead

Returns the raw block markup for a single post, page, or synced pattern by ID or URL (including drafts) so an agent can read exactly what it is about to test, back up, or redesign. Guarded by per-post edit permission.

Ask the AI

Show me the raw blocks for the homepage hero.

ParameterTypeRequiredDescription
idintegerNoPost, page, or block ID. Takes precedence over url when both are given.
urlstringNoPost URL, used to resolve the content when no id is given.

Example Request

json
{
  "name": "accelerate/get-content",
  "arguments": { "url": "https://your-site.com/pricing/" }
}

accelerate/get-media

DiscoveryRead

Lists media-library attachments (URL, alt text, dimensions, and MIME type) with optional search and MIME-type filters and pagination, so an agent can ingest a site’s real imagery rather than guessing from block markup.

ParameterTypeRequiredDescription
limitintegerNoItems per page, 1–100. Defaults to 20.
offsetintegerNoNumber of items to skip for pagination. Defaults to 0.
searchstringNoOptional text filter on title or filename.
mime_typestringNoOptional MIME-type filter, such as image/jpeg or application/pdf.