Key concepts
The few terms that the app, the reports and the API share, and how they relate.
The app, the reports and the API use the same words with the same meaning. Learn these and every other page reads easily.
Site
A website you want to audit, identified by its origin (https://www.example.com). A site has a name, the address an audit reads, an optional repeating schedule, and the result of its latest audit. You can keep 10 sites per account.
In the API a site is a Site object with an id that starts with site_.
Audit run
One audit of one site at one moment. A run moves through these statuses:
| Status | Meaning |
|---|---|
queued |
Waiting in the queue. A scheduled run waits here until its time. |
running |
Being worked on now. |
succeeded |
Finished; every check produced a result. |
partial |
Finished; some checks could not be measured. The rest are valid. |
failed |
Did not finish. errorCode says why, for example url_policy_denied. |
queued and running are temporary; the other three are final. A run id starts with run_.
A run records how it started (triggerType): manual from the app, api from a key, scheduled from a schedule.
Profile
The set of checks a run uses. The live profile is the passive audit, recorded on every run as profileVersion. When the profile changes, old runs keep the version they ran with, so results stay comparable.
Area
Checks are grouped into areas. The contract knows these: availability, security, email, performance, marketing, accessibility, privacy and technology. The report shows one tile per area.
Observation
The result of one check in one run: the area, the check id, a status, a severity, a confidence and a one-line summary of what was seen. In the app an observation is one row of the report.
Observation status
| Status | In the app | Meaning |
|---|---|---|
pass |
Passed | The check passed. |
fail |
To fix | The check failed. |
no_data |
No data | Nothing to check. Not a failure. |
not_tested |
Not tested | This profile cannot see it. Not a failure. |
failed |
Not measured | The measurement failed, not the site. |
no_data and not_tested are never counted as failures, in any output.
Severity and confidence
Severity is high, medium, low or info and says how much a failed check matters. Confidence (high, medium, low) says how sure the observation is, given what a passive read can see.
Schedule
A site can repeat its audit weekly or monthly, or not at all (none). The first repeated run starts one period after you set it. A single later run is not a schedule: it is a run with a start time (scheduledFor), up to 90 days ahead.
API key
A secret that lets a script act as your account, limited to the scopes you choose. See API keys and Authentication and scopes.
Problem report
A message to the Metricevo team about a bug, a wrong result or a question, optionally tied to one audit run. You can follow its status: open, in_progress, closed.