Kilotest

Blocks and waits for completion (or failure or 5-minute timeout) of a requested job. Call this tool if you requested a new test or retest with immediate review and you want to block the workflow until the test is complete. Requires a reportID from requestNewTest or requestRetest. If you choose not to wait or the wait times out, call listIssues with the reportID later to check whether the results are available.

Gets an overview of Kilotest, explaining that it is a front-end quality testing service for the web and describing its tools and the structure of its test results. Call this before any other tool when the tools and workflow are unclear.

Gets an entire report as JSON. Call this tool if you want the full report for analysis outside of Kilotest. Requires a reportID from listPages or listIssues. The listIssues tool discloses the size of a report (typically 0.5 to 2 MB) before you decide whether to download it. If you want only selected results, drill down with listIssues, listViolators, and/or listDiagnoses instead.

Lists the diagnoses by rule engines for one violator of rules belonging to one issue in one report, each with a description and severity. Call this after listViolators to continue drilling down. Requires a reportID, issueID, and catalogIndex (element identifier) from listViolators. After this tool, if you want to question a diagnosis, call questionDiagnosis. If you want the entire JSON report for analysis outside Kilotest, call getReport instead of listIssues, listViolators, or this tool.

Lists the issues reported in one report, each with an identifier, summary, and priority, sorted by summary. Also lists all the reports about the same page chronologically. Call this tool after listPages to begin getting test results about a page (optionally continuing to drill down with listViolators and listDiagnoses), or to switch to an earlier or later report. Also call this tool after requestNewTest or requestRetest to learn whether the requested test results are available. Requires a reportID from listPages, the list of reports, requestNewTest, or requestRetest. After this tool, if you want to know which elements violated rules belonging to one of the issues, call listViolators. If you want the entire JSON report for analysis outside Kilotest, call getReport instead of this tool, listViolators, or listDiagnoses.

Lists the tested pages of one site, each with a page identifier, name, and URL, an array of the sites it belongs to (currently always 1), and the latest report identifier, sorted by page name. Call this after listSites to learn which pages of a site have been tested. Requires a siteID from listSites. After this tool, if you want to know which issues were reported, call listIssues. If you want a different page of the site tested, call requestNewTest.

Lists all tested sites, alphabetically by name, each with its identifier and name and the count of its tested pages. Call this tool first to learn whether a site has been tested. If it has, call listPages to learn which of its pages have been tested. If the site has not been tested and you want it tested, call requestNewTest.

Lists the elements of one page that were reported as violators of rules belonging to one issue in one report, each with its identifier (catalogIndex), tag name, inner text, and count of reporting rule engines, sorted by identifier. Call this tool after listIssues to continue drilling down. Requires an issueID and a reportID from listIssues. After this tool, if you want to know the diagnoses for a specific violator, call listDiagnoses. If you want the entire JSON report for analysis outside Kilotest, call getReport instead of listIssues, this tool, or listDiagnoses.

Records and submits to management a disagreement with or doubts about one diagnosis reported by the listDiagnoses tool. Call this tool to question a diagnosis or its issue classification. Requires a reportID, issueID, catalogIndex (element identifier), engineID, and ruleID from listDiagnoses, and a reason explaining your concern. If your objection is more general than a single diagnosis, call requestFeature instead.

Requests a new feature or an improvement to an existing one. Requires a feature parameter describing the requested feature or improvement. If you want to question a diagnosis, call questionDiagnosis instead.

Requests that a not yet tested page be tested. Call this tool after listSites or listPages. Requires either a siteID from listSites if the site has been tested, or a siteName, namely the brand or organization name of the site, like the names shown by listSites, but distinct from the name of any tested site when lower-cased and stripped of punctuation and white space. Also requires a pageName describing the purpose of the page, such as home, about, contact, shop, blog, subscribe, sign up, or Benefactors Club; a URL; and a reason justifying the request. If the site has been tested, the pageName must be distinct from the names of all other tested pages of the site shown by listPages. Permits an optional boolean reviewImmediately parameter. If it is true, the response will notify you of approval (with a report identifier, and with an opportunity to call awaitTest and wait for, typically, 2 to 4 minutes), rejection, or queuing for management review. If reviewImmediately is false or omitted, the request will be queued for management review (approval and test completion typically take an hour to a day). If the page has already been tested, call requestRetest instead.

Requests that a page be tested again. Call this tool if you have reason to consider the latest report about the page outdated. Requires a reportID from listPages or listIssues and a reason justifying the request. Permits an optional boolean reviewImmediately parameter. If it is true, the response will notify you of approval (with a report identifier, and with an opportunity to call awaitTest and wait for, typically, 2 to 4 minutes), rejection, or queuing for management review. If reviewImmediately is false or omitted, the request will be queued for management review (approval and test completion typically take an hour to a day). If the page has not yet been tested, call requestNewTest instead.