index365usa index365 read report
Read a scan's score and findings, and explain them. Read-only, and no ids needed:
every command defaults to the latest completed run and echoes what it resolved
(yoursite.com · scanned 2m ago). A domain narrows to that site's latest run.
Score first
index365 report
On a terminal this renders the score card; piped or with --json it prints the
compact agent-ready JSON context (score, severity counts, top findings, exec
summary). That JSON is bounded by construction, so it is safe to read directly.
Then findings
index365 findings # table from the latest run
index365 findings get 1 # first finding in full, plus a fix prompt
Filter the table with --severity critical (then high) or
--marketing-category <c> (seo, content, aeo, analytics_tracking,
conversion, trust_proof). --stage <s> still filters legacy runs completed
before the category model; it matches nothing on a current run. findings get <n> takes the ordinal straight from the
table; each finding also has a stable findingId that stays constant across
re-reads, which is what you match on when comparing runs.
Large reports: save, then read selectively
Only when you need every finding at once (full triage, cross-run diffing), save the
full report to the git-ignored .index365/ dir and read it with jq / grep:
index365 report --save # writes the full report into .index365/ by default
Never paste a whole saved report into context. Summarize the score, the severity counts, and the top findings.
History
index365 results lists your scans newest first; index365 results yoursite.com is
that site's score over time, the fastest way to show whether a fix moved the number.
Next
To act on the findings: index365-triage-findings (plan), index365-apply-fix (one fix), or index365-audit-and-fix (the full loop).