Get the AI readiness report
const url = 'https://api.overwatch.weareheavy.dev/v1/sites/2489E9AD-2EE2-8E00-8EC9-32D5F69181C0/ai-report';const options = {method: 'GET', headers: {'x-api-key': '<x-api-key>'}};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request GET \ --url https://api.overwatch.weareheavy.dev/v1/sites/2489E9AD-2EE2-8E00-8EC9-32D5F69181C0/ai-report \ --header 'x-api-key: <x-api-key>'The AI readiness report, derived live from the latest crawl that
captured AI signals: findings grouped by rule (ranked by severity, then
by recoverable score points), a worst-first per-page view, and a summary
with the 0-100 score, a verdict and the per-bot robots.txt access
table.
The score measures whether AI assistants can read and cite the site,
not whether they currently mention it. Training-bot blocks and
llms.txt are reported as notes and never scored.
Issues are derived at read time from per-page snapshots, so changes to
the site’s excludePaths globs apply retroactively. The score frozen
on the run row feeds the trend; the report’s live score can differ
briefly after a settings change.
Authorizations
Section titled “Authorizations”Parameters
Section titled “ Parameters ”Path Parameters
Section titled “Path Parameters”The site’s UUID.
Responses
Section titled “ Responses ”The latest AI-capturing run, its findings, and summary.
object
A single crawl run and its summary metrics.
object
Epoch ms the run started.
Epoch ms the run finished, or null while running.
Distinct broken link targets found in the run.
Distinct source pages that had at least one broken link.
Synthesized 0–100 accessibility score, or null when the run didn’t scan.
Confirmed violations with critical impact.
Internal pages the accessibility scan covered in this run.
Distinct confirmed (page, rule) accessibility violations this run (0 for crawl runs).
Violations present in the previous accessibility run but gone in this one (the “fixed” trend).
Synthesized 0-100 SEO score, or null when the run didn’t extract SEO data.
Pages an SEO snapshot was captured for (at most pages_crawled).
Derived SEO issues at the time the run finalized.
Synthesized 0-100 AI readiness score, or null when the run didn’t capture AI signals.
Pages an AI snapshot was captured for (at most pages_crawled).
Derived AI readiness issues at the time the run finalized.
JSON-encoded site-wide facts captured at seed (llmsTxt,
cloudflare, trainingBlocked), or null. The /ai-report
summary.notes is the parsed form.
SEO issues opened this run. null on starting-point and pre-lifecycle runs.
SEO issues fixed this run (the page was crawled and the rule no longer fires).
SEO issues closed because their page is no longer crawled. Not counted as fixed.
Fixed SEO issues that came back this run.
Pages that answered with a bot-protection challenge instead of their content. null on runs from before crawl-health tracking.
Pages that refused access (401/403, or still rate-limited after every retry). Never reported as broken. null on runs from before tracking.
Pages that were rate-limited at least once (the crawl slowed down and retried). null on runs from before tracking.
Failure reason if state is failed, otherwise null.
crawl for a broken-link crawl, a11y for a standalone accessibility scan.
Findings grouped per rule, most severe and most impactful first.
object
Stable rule id (e.g. ai-search-bot-blocked, ai-content-js-only).
One-paragraph editor-facing remediation guidance.
true for a rule evaluated once for the whole site
(ai-no-organization). It has no examples and
counts against every scanned page.
Pages affected by this rule (all scanned pages when siteWide).
Score points recoverable by fixing this rule site-wide.
Affected pages (capped); empty for site-wide rules.
object
Every page with at least one finding, ranked worst-first. Site-wide findings are not listed here.
object
The page’s own 0-100 score.
object
object
object
blocked when a major AI search bot is blocked on
at least a quarter of the scanned pages;
otherwise hard-to-read when any critical or
serious finding exists; else readable.
Per-bot robots.txt access rolled up across the
scanned pages, in catalog order. search bots are
scored; training bots are noted only.
object
One AI crawler’s robots.txt access across the scanned pages.
object
The robots.txt product token (e.g. OAI-SearchBot).
allowed = no page blocked, partial = some pages blocked, blocked = every page blocked.
Extra context for the bot, if any (e.g. Google-Extended also affects Gemini grounding).
One AI crawler’s robots.txt access across the scanned pages.
object
The robots.txt product token (e.g. OAI-SearchBot).
allowed = no page blocked, partial = some pages blocked, blocked = every page blocked.
Extra context for the bot, if any (e.g. Google-Extended also affects Gemini grounding).
Pages this crawl couldn’t read the way a visitor would, from the run’s
pages_challenged / pages_blocked / pages_rate_limited. null when
every page came through, or on runs from before tracking. Challenged and
blocked pages are missing from the report’s findings.
object
Example
{ "run": { "state": "running", "kind": "crawl" }, "byRule": [ { "severity": "critical" } ], "byPage": [ { "findings": [ { "severity": "critical" } ] } ], "summary": { "verdict": { "status": "readable" }, "bots": { "search": [ { "status": "allowed" } ], "training": [ { "status": "allowed" } ] } }}Missing or invalid credentials.
object
A human-readable error message.
Example
{ "error": "invalid api key"}The site was not found, the AI readiness app is not enabled, or no completed crawl has captured AI signals yet.
object
A human-readable error message.
Examples
App not installed
{ "error": "AI readiness is not enabled for this site"}No crawl has captured AI signals yet
{ "error": "no runs yet"}