API reference

Scanner

Start scans and read what they found: cookies, requests, findings, reports.

PATCH/v1/cookies/{id}
GET/v1/domains/{domainId}/cookies
GET/v1/domains/{domainId}/declaration
PUT/v1/domains/{domainId}/deep-scan
GET/v1/domains/{domainId}/enforcement
GET/v1/domains/{domainId}/enforcement.pdf
GET/v1/domains/{domainId}/enforcement/history
GET/v1/domains/{domainId}/scan-settings
GET/v1/domains/{domainId}/scans
POST/v1/domains/{domainId}/scans
GET/v1/enforcement/summary
GET/v1/scans/{id}
GET/v1/scans/{id}/findings

PATCH/v1/cookies/{id}

Override a cookie’s category

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
id required	path	string

Request body · OverrideCookieDto
Field	Type	What it is
category required	"unknown" \| "marketing" \| "preferences" \| "analytics" \| "necessary"
vendor	string	The vendor, if you know better than the scan. Optional; an empty string clears it.

Responses
Status	Body	When
200	CookieDto

200 body · CookieDto
Field	Type	What it is
id required	string
name required	string
domainScope required	string
party required	"first" \| "third"
category required	"unknown" \| "marketing" \| "preferences" \| "analytics" \| "necessary"
vendor required	string \| null
purpose required	string \| null
expiry required	string \| null
overridden required	boolean
underReview required	boolean	True when the category is unknown because a suggestion is waiting for a person to confirm it, rather than because nothing is known. A suggestion below the confidence threshold is never written into a cookie declaration as settled.

GET/v1/domains/{domainId}/cookies

Cookies from a domain’s latest completed scan (filterable)

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string
category	query	"necessary" \| "analytics" \| "marketing" \| "preferences" \| "unknown"
party	query	"first" \| "third"
q	query	string	Case-insensitive cookie-name search.

Responses
Status	Body	When
200	CookieDto[]

200 body · CookieDto[]
Field	Type	What it is
id required	string
name required	string
domainScope required	string
party required	"first" \| "third"
category required	"unknown" \| "marketing" \| "preferences" \| "analytics" \| "necessary"
vendor required	string \| null
purpose required	string \| null
expiry required	string \| null
overridden required	boolean
underReview required	boolean	True when the category is unknown because a suggestion is waiting for a person to confirm it, rather than because nothing is known. A suggestion below the confidence threshold is never written into a cookie declaration as settled.

GET/v1/domains/{domainId}/declaration

Cookie declaration (JSON by default, ?format=html)

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string
format	query	"json" \| "html"

Responses
Status	Body	When
200	DeclarationDto

200 body · DeclarationDto
Field	Type	What it is
generatedAt required	string
totalCookies required	number
categories required	DeclarationCategoryDto[]
categories[].category required	"unknown" \| "marketing" \| "preferences" \| "analytics" \| "necessary"
categories[].count required	number
categories[].cookies required	DeclarationCookieDto[]

PUT/v1/domains/{domainId}/deep-scan

Nominate this domain for the monthly deep scan

Agency and above. The plan’s allowance is enforced here, not in the panel. Turning one off is always permitted, even while over the allowance — a plan downgrade can leave a workspace above its own limit.

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string

Request body · SetDeepScanDto
Field	Type	What it is
enabled required	boolean

Responses
Status	Body	When
200	DomainScanSettingsDto

200 body · DomainScanSettingsDto
Field	Type	What it is
domainId required	string
hostname required	string
cadence required	"weekly" \| "monthly"	How often a routine scan runs for this plan.
routinePages required	number	Pages a routine run may visit.
deepPages required	number	Pages a deep run may visit. Zero on a plan with no deep scan.
routineEstimatedMs required	number	The longest a routine scan of this domain can take, in milliseconds, and the same for a deep scan where the plan has one. A ceiling, not a prediction.
deepEstimatedMs required	number	Zero where `deepPages` is zero.
deepDomainsAllowed required	number	How many domains this workspace may nominate. Zero below Agency.
deepDomainsUsed required	number	How many it has nominated, including this one.
deepScan required	boolean	Is this domain one of them?
nextRunAt required	string \| null	Null until the domain is verified — an unverified domain is never crawled.
nextRunBlockedBy required	"unverified" \| "in_flight" \| null	Why there is no next run, when there is none.
nextRunDeep required	boolean	Whether the next scheduled run is the monthly deep scan. It takes the place of that period's routine run; between deep runs the routine schedule carries on.
lastRunAt required	string \| null
lastStatus required	"queued" \| "failed" \| "blocked" \| "running" \| "completed" \| null
lastVerdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined" \| null
lastBlockedReason required	string \| null	The reason the site turned us away, verbatim, for the panel to explain — taken from the latest scan that reached the site, so a later failure on our side does not hide it. Null when that scan got through.
lastRegion required	string \| null
scannerUserAgent required	string	How the crawler identifies itself — what a WAF rule allows.
egressAddresses required	string[]	The addresses the scanner connects from, for a firewall rule. Empty when they have not been published.
scannerInfoUrl required	string	The public page that explains the crawler to a site owner.
routineAllowance required	ScanAllowanceDto	The routine allowance, which Scan now spends.
routineAllowance.total required	number	Pages the plan gives per period.
routineAllowance.used required	number	Pages spent in the current window.
routineAllowance.remaining required	number	What a scan started now may visit; zero when spent.
routineAllowance.nextAvailableAt required	string \| null	When a scan becomes possible again, ISO-8601; null unless spent.
routineAllowance.windowDays required	number	The rolling window, in days: 30 for monthly, 7 for weekly.
deepAllowance required	ScanAllowanceDto \| null	The deep-scan allowance, or null where the plan has no deep scans.
deepAllowance.total required	number	Pages the plan gives per period.
deepAllowance.used required	number	Pages spent in the current window.
deepAllowance.remaining required	number	What a scan started now may visit; zero when spent.
deepAllowance.nextAvailableAt required	string \| null	When a scan becomes possible again, ISO-8601; null unless spent.
deepAllowance.windowDays required	number	The rolling window, in days: 30 for monthly, 7 for weekly.

GET/v1/domains/{domainId}/enforcement

The enforcement report for a domain

The latest version, or a named one with ?version=. A report is a fixed snapshot of what one scan proved, plus the diff against the previous version: what appeared, what was fixed, and what came back after being fixed — which is a different conversation from what is merely new.

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string
version	query	number

Responses
Status	Body	When
200	EnforcementReportDto

200 body · EnforcementReportDto
Field	Type	What it is
id required	string
domainId required	string
hostname required	string
scanId required	string
version required	number	Per domain, 1..n. What a customer cites.
verdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined"
critical required	number
warnings required	number
passes required	number	Findings that are good news: correct gating, compliant denied pings.
pagesVisited required	number
pageBudget required	number \| null	Pages the plan allowed. Null when the scan predates budgets.
region required	string \| null
added required	string[]	Fingerprints new since the previous version — never seen before.
fixed required	string[]	In the previous version, gone now.
regressed required	string[]	Fixed once and back again, which is a different conversation from `added`.
scannedAt required	string \| null	When the scan ran, as against when the report row was written.
createdAt required	string
findings required	ScanFindingDto[]
findings[].id required	string
findings[].type required	"consent_state_mismatch" \| "leak" \| "failed_to_fire" \| "fired_correctly" \| "cookieless_ping"
findings[].severity required	"info" \| "warning" \| "critical"
findings[].vendorKey required	string \| null	Catalogue key where the host matched a known vendor, else null.
findings[].vendorName required	string
findings[].host required	string
findings[].pageUrl required	string
findings[].passKind required	"accept" \| "reject" \| "no_interaction"	Which pass proves it — before any choice, after reject, or after accept.
findings[].title required	string
findings[].detail required	string
findings[].fingerprint required	string	Stable across scans for the same problem on the same page.
findings[].evidence required	object[]
findings[].explanation required	ScanFindingExplanationDto \| null	An AI-written explanation of the same finding, or null. It is labelled as generated wherever it is shown, and nothing in the verdict, the counts or the evidence comes from it.

GET/v1/domains/{domainId}/enforcement.pdf

The enforcement report as a PDF an agency can forward

Cover, summary, every finding with the request lines it rests on, and a methodology page — because the reader was not in the room and their first question is how we know. Contains no visitor data: the recorder stores cookie names, never values, and this quotes only those.

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string
version	query	number

Responses
Status	Body	When
200	—

GET/v1/domains/{domainId}/enforcement/history

Every report version for a domain, newest first

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string

Responses
Status	Body	When
200	EnforcementHistoryEntryDto[]

200 body · EnforcementHistoryEntryDto[]
Field	Type	What it is
id required	string
version required	number
scanId required	string
verdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined"
critical required	number
warnings required	number
added required	number
fixed required	number
regressed required	number
createdAt required	string

GET/v1/domains/{domainId}/scan-settings

When this domain is next scanned, how many pages it gets, and why not

Worked out from the plan and the schedule. nextRunAt is an estimate: due scans are started every hour, so the run happens at or after that time.

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string

Responses
Status	Body	When
200	DomainScanSettingsDto

200 body · DomainScanSettingsDto
Field	Type	What it is
domainId required	string
hostname required	string
cadence required	"weekly" \| "monthly"	How often a routine scan runs for this plan.
routinePages required	number	Pages a routine run may visit.
deepPages required	number	Pages a deep run may visit. Zero on a plan with no deep scan.
routineEstimatedMs required	number	The longest a routine scan of this domain can take, in milliseconds, and the same for a deep scan where the plan has one. A ceiling, not a prediction.
deepEstimatedMs required	number	Zero where `deepPages` is zero.
deepDomainsAllowed required	number	How many domains this workspace may nominate. Zero below Agency.
deepDomainsUsed required	number	How many it has nominated, including this one.
deepScan required	boolean	Is this domain one of them?
nextRunAt required	string \| null	Null until the domain is verified — an unverified domain is never crawled.
nextRunBlockedBy required	"unverified" \| "in_flight" \| null	Why there is no next run, when there is none.
nextRunDeep required	boolean	Whether the next scheduled run is the monthly deep scan. It takes the place of that period's routine run; between deep runs the routine schedule carries on.
lastRunAt required	string \| null
lastStatus required	"queued" \| "failed" \| "blocked" \| "running" \| "completed" \| null
lastVerdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined" \| null
lastBlockedReason required	string \| null	The reason the site turned us away, verbatim, for the panel to explain — taken from the latest scan that reached the site, so a later failure on our side does not hide it. Null when that scan got through.
lastRegion required	string \| null
scannerUserAgent required	string	How the crawler identifies itself — what a WAF rule allows.
egressAddresses required	string[]	The addresses the scanner connects from, for a firewall rule. Empty when they have not been published.
scannerInfoUrl required	string	The public page that explains the crawler to a site owner.
routineAllowance required	ScanAllowanceDto	The routine allowance, which Scan now spends.
routineAllowance.total required	number	Pages the plan gives per period.
routineAllowance.used required	number	Pages spent in the current window.
routineAllowance.remaining required	number	What a scan started now may visit; zero when spent.
routineAllowance.nextAvailableAt required	string \| null	When a scan becomes possible again, ISO-8601; null unless spent.
routineAllowance.windowDays required	number	The rolling window, in days: 30 for monthly, 7 for weekly.
deepAllowance required	ScanAllowanceDto \| null	The deep-scan allowance, or null where the plan has no deep scans.
deepAllowance.total required	number	Pages the plan gives per period.
deepAllowance.used required	number	Pages spent in the current window.
deepAllowance.remaining required	number	What a scan started now may visit; zero when spent.
deepAllowance.nextAvailableAt required	string \| null	When a scan becomes possible again, ISO-8601; null unless spent.
deepAllowance.windowDays required	number	The rolling window, in days: 30 for monthly, 7 for weekly.

GET/v1/domains/{domainId}/scans

A domain’s recent scans, with the pages each one visited

Newest first, capped at 50. Each scan carries its three passes per page, how many requests and storage entries each pass saw, and whether the pass could operate the site’s consent controls at all.

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string

Responses
Status	Body	When
200	ScanDetailDto[]

200 body · ScanDetailDto[]
Field	Type	What it is
id required	string
domainId required	string
status required	"queued" \| "failed" \| "blocked" \| "running" \| "completed"
pagesCrawled required	number
pageBudget required	number \| null	Pages this run was allowed to visit, from the plan as it was when the scan was queued. Sent so a report can say "25 of 25 pages" rather than leaving a reader to assume it saw the whole site — which is the assumption that makes a clean result on a partial crawl dangerous.
pagesPlanned required	number \| null	Pages this run actually set out to visit — discovery's result, capped by the budget. The denominator for progress, and null on scans that predate it. Separate from `pageBudget` because the budget is what the plan allows: a five-page site on a fifty-page plan is at 100% after five, not 10%.
estimatedMs required	number	The longest this scan can take, in milliseconds. A ceiling, not a prediction: most scans finish in well under a third of it.
deep required	boolean	A deep scan: the monthly, far larger look at a nominated domain.
region required	string \| null	The region the scan ran from. Null until the scan has started.
verdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined"	What the scan concluded. The findings endpoint has the detail.
error required	string \| null
startedAt required	string \| null
finishedAt required	string \| null
createdAt required	string
passes required	ScanPassSummaryDto[]
passes[].pageUrl required	string
passes[].kind required	"accept" \| "reject" \| "no_interaction"
passes[].reachedDecision required	boolean	Whether the scanner could operate the site's consent controls. Read this first: a pass that did not reach a decision proves nothing about the site.
passes[].reason required	string \| null	Why not, when not: "no reject control found", "blocked by a bot challenge".
passes[].requests required	number
passes[].storage required	number
passes[].ms required	number
pagesVisited required	number	Distinct pages visited, across the three passes.
pagesUndetermined required	number	Pages where at least one pass could not be operated.
criticalFindings required	number
warningFindings required	number

POST/v1/domains/{domainId}/scans

Queue a cookie scan for a domain

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
domainId required	path	string

Responses
Status	Body	When
201	ScanDto

201 body · ScanDto
Field	Type	What it is
id required	string
domainId required	string
status required	"queued" \| "failed" \| "blocked" \| "running" \| "completed"
pagesCrawled required	number
pageBudget required	number \| null	Pages this run was allowed to visit, from the plan as it was when the scan was queued. Sent so a report can say "25 of 25 pages" rather than leaving a reader to assume it saw the whole site — which is the assumption that makes a clean result on a partial crawl dangerous.
pagesPlanned required	number \| null	Pages this run actually set out to visit — discovery's result, capped by the budget. The denominator for progress, and null on scans that predate it. Separate from `pageBudget` because the budget is what the plan allows: a five-page site on a fifty-page plan is at 100% after five, not 10%.
estimatedMs required	number	The longest this scan can take, in milliseconds. A ceiling, not a prediction: most scans finish in well under a third of it.
deep required	boolean	A deep scan: the monthly, far larger look at a nominated domain.
region required	string \| null	The region the scan ran from. Null until the scan has started.
verdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined"	What the scan concluded. The findings endpoint has the detail.
error required	string \| null
startedAt required	string \| null
finishedAt required	string \| null
createdAt required	string

GET/v1/enforcement/summary

What moved across the workspace this week

Sums the diffs of reports written in the window rather than counting findings open now: the question is what happened while you were not looking, and the compliance score already answers the other one.

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Responses
Status	Body	When
200	EnforcementWeekDto

200 body · EnforcementWeekDto
Field	Type	What it is
windowDays required	number	How many days the window covers.
added required	number	Problems that appeared, across every domain.
fixed required	number	Problems that went away.
regressed required	number	Problems that were fixed and came back.
domainsWithCritical required	number	Domains with at least one critical finding in their latest report.
domainsReported required	number	Domains that have a report at all — the denominator for the above.
latest required	EnforcementLatestDto \| null	The report to link to: the newest one in the window that found something new or returning, otherwise the newest report of all. Null when no domain has a report.
latest.domainId required	string
latest.hostname required	string
latest.version required	number

GET/v1/scans/{id}

Get a scan, with the pages it visited and what each pass did

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
id required	path	string

Responses
Status	Body	When
200	ScanDetailDto

200 body · ScanDetailDto
Field	Type	What it is
id required	string
domainId required	string
status required	"queued" \| "failed" \| "blocked" \| "running" \| "completed"
pagesCrawled required	number
pageBudget required	number \| null	Pages this run was allowed to visit, from the plan as it was when the scan was queued. Sent so a report can say "25 of 25 pages" rather than leaving a reader to assume it saw the whole site — which is the assumption that makes a clean result on a partial crawl dangerous.
pagesPlanned required	number \| null	Pages this run actually set out to visit — discovery's result, capped by the budget. The denominator for progress, and null on scans that predate it. Separate from `pageBudget` because the budget is what the plan allows: a five-page site on a fifty-page plan is at 100% after five, not 10%.
estimatedMs required	number	The longest this scan can take, in milliseconds. A ceiling, not a prediction: most scans finish in well under a third of it.
deep required	boolean	A deep scan: the monthly, far larger look at a nominated domain.
region required	string \| null	The region the scan ran from. Null until the scan has started.
verdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined"	What the scan concluded. The findings endpoint has the detail.
error required	string \| null
startedAt required	string \| null
finishedAt required	string \| null
createdAt required	string
passes required	ScanPassSummaryDto[]
passes[].pageUrl required	string
passes[].kind required	"accept" \| "reject" \| "no_interaction"
passes[].reachedDecision required	boolean	Whether the scanner could operate the site's consent controls. Read this first: a pass that did not reach a decision proves nothing about the site.
passes[].reason required	string \| null	Why not, when not: "no reject control found", "blocked by a bot challenge".
passes[].requests required	number
passes[].storage required	number
passes[].ms required	number
pagesVisited required	number	Distinct pages visited, across the three passes.
pagesUndetermined required	number	Pages where at least one pass could not be operated.
criticalFindings required	number
warningFindings required	number

GET/v1/scans/{id}/findings

What a scan proved about enforcement

The deterministic findings for a scan: leaks, consent-state mismatches, vendors that fired correctly and vendors that never fired. verdict is the part to read first — undetermined means the crawler reached the site but could not operate its consent controls, which is emphatically not "clean".

Needs: A signed-in panel session, sent as Authorization: Bearer <access token>.

Parameters
Name	Where	Type	What it is
id required	path	string

Responses
Status	Body	When
200	ScanFindingsDto

200 body · ScanFindingsDto
Field	Type	What it is
scanId required	string
verdict required	"blocked" \| "not_run" \| "assessed" \| "undetermined"
assessedAt required	string \| null	When the assertions last ran, or null when they have not.
critical required	number
warnings required	number
passes required	number	Findings that are good news: correct gating and compliant denied pings.
findings required	ScanFindingDto[]
findings[].id required	string
findings[].type required	"consent_state_mismatch" \| "leak" \| "failed_to_fire" \| "fired_correctly" \| "cookieless_ping"
findings[].severity required	"info" \| "warning" \| "critical"
findings[].vendorKey required	string \| null	Catalogue key where the host matched a known vendor, else null.
findings[].vendorName required	string
findings[].host required	string
findings[].pageUrl required	string
findings[].passKind required	"accept" \| "reject" \| "no_interaction"	Which pass proves it — before any choice, after reject, or after accept.
findings[].title required	string
findings[].detail required	string
findings[].fingerprint required	string	Stable across scans for the same problem on the same page.
findings[].evidence required	object[]
findings[].explanation required	ScanFindingExplanationDto \| null	An AI-written explanation of the same finding, or null. It is labelled as generated wherever it is shown, and nothing in the verdict, the counts or the evidence comes from it.

Next groupAnalytics