Insights · Pro, Pro Team & Enterprise
“What changed in the payments module this month?” is a simple question with an expensive answer: a Slack interruption, a trawl through pull requests, an afternoon of git archaeology. The knowledge exists — it’s just spread across every review your team ever shipped.
PRGuard already audits every pull request and push. Insights makes that history conversational: ask in plain language, get an answer grounded in the audit trail — with a citation for every claim, linking straight to the audit it came from.
One question, answered from the audit record — not from anyone’s memory.
Illustrative example
“Which findings from last month are still unfixed?”
Two findings remain open from July’s audits. The refund flow is still carrying an error-severity SQL-handling finding from PR #151 — assigned, not yet fixed [1]. A warning about missing rate limiting on the invite endpoint is unassigned [2]. Everything else flagged in July was either fixed (six findings, three by auto-fix PRs) or dismissed with a reason.
Sources: [1] PR #151 audit · [2] PR #158 audit — each links to the full record.
Insights is not a chatbot with opinions. It is a narrow set of read-only queries over your own audit records, narrated.
1 · Only your audit trail
Answers draw exclusively on the audits PRGuard ran for your organisation — the changed files, verdicts, findings, and fix statuses already in your record. The AI chooses search filters; it never sees anything beyond what those org-scoped queries return.
2 · Real paths, not guesses
Ask about “payments” and Insights first looks up which files actually changed in the window, then filters by the real paths it finds — so a fuzzy module name resolves to your repository’s structure, not a guess at it.
3 · Flagged vs still open
Every finding carries its lifecycle — open, in progress, fixed, or dismissed — so answers distinguish “was flagged in review” from “is still a problem today”. That distinction is usually the whole point of asking.
4 · Citations are enforced
Each claim cites the audit it came from, and citations are validated server-side: a reference that doesn’t match an audit retrieved for that answer is stripped. If the history shows nothing for the window, the answer says so.
Release notes and status updates without interrupting the team.
“What changed in checkout in the last two weeks?”
Attribution joined to audit outcomes, not just git blame.
“Who has been changing our auth code, and what did audits flag?”
Cited answers that point at the evidence record itself.
“Summarise everything that failed an audit this quarter and why.”
Insights answers from the changes PRGuard audited: pull requests and pushes on your connected repositories, from the day you connected them. It does not see commits that bypassed review on unconnected branches, or history from before PRGuard was installed — and when a window is only partially covered, the answer says so rather than papering over the gap.
For a governance tool, that boundary is the feature: every statement traces to an audit record you can open, audit, and export.
Connect a repository, let the audits accrue, and ask.