Beta

Your PRs remember
what broke before.

When a pull request touches a file with a troubled past, Pullminder surfaces that history in the review comment. For higher-risk PRs, it also appears in the AI brief. No configuration needed after opt-in.

Enable in dashboard Read the guide

How it works

Three signals, one timeline

Collective Memory tracks three signals across every repository that opts in. Each signal is computed from pull request and commit history that Pullminder already has access to.

Revert and hotfix history

Files that have been repeatedly involved in reverts, hotfixes, or post-merge patches. When a PR touches one of these files, Pullminder notes the pattern so reviewers know the area has been trouble before.

Churn

Files that are frequently re-patched shortly after merge — a signal that previous fixes did not stick. High churn correlates with under-specified requirements and incomplete review.

Recurring findings and bus factor

Which finding categories keep surfacing in a repo, and how concentrated authorship is. A single-author file with repeated security findings tells a different story than a well-distributed file with occasional style nits.

Memory is surfaced in the PR comment, the AI brief for higher-risk pull requests, and on the dashboard analytics page with per-repo trend charts and an org-wide troubled-files view. It is display-only — Collective Memory never changes your risk score.

Privacy

Opt-in, off by default, purgeable at any time

Collective Memory is a Beta feature built with a minimal-data posture. It stores file-path metadata and aggregate counts only — no source code, no diff content, and no comment text.

Per-repository opt-in

Off by default for every repository. Enable it from the dashboard repo settings when you are ready.

File-path metadata only

No code is stored. Pullminder records file paths, event types, and aggregate counts. That is it.

Disabling purges memory

Turn off Collective Memory for a repository and all associated data is purged immediately and irreversibly.

GDPR erasure supported

Author-scoped data scrub and full-repository purge are both available. Contact support or use the dashboard controls.

FAQ

Questions you probably have

What data does Collective Memory store?
File-path metadata and aggregate counts only. Pullminder records which files were involved in reverts, hotfixes, churn events, or past findings, plus summary statistics (frequency, recency, author count). No source code, no diff content, no comment text is stored.
How do I turn it on or off?
Collective Memory is off by default. Enable it per-repository from your dashboard's repo settings. Disabling a repository immediately purges all of its stored memory.
Does Collective Memory affect the risk score?
No. Collective Memory is display-only. It adds a history section to the PR comment and, for higher-risk PRs, the AI brief, but it never changes the risk score or the pass/fail threshold.
Does it work with private repositories?
Yes. Collective Memory works with private repositories on all plans. The same opt-in and purge controls apply regardless of repository visibility.
How long is the data retained?
Memory is retained for the lifetime of the repository's opt-in. If you disable Collective Memory for a repo, all associated data is purged immediately. GDPR erasure requests are supported — author-scoped scrub and full-repo purge are both available.

Give your PRs a memory

Opt in from the dashboard. One toggle per repository. No config files, no CLI flags, no code changes.

Enable Collective Memory Read the docs