kalpha
kalpha's work was selected for the XERJ Corpus Hub, the open reference memory AI coding agents search before they write code. When an agent meets the same problem, it gets this implementation back as a worked example, with the author and the repository attached.
Totals across every corpus @gguatit appears in; a commit selected for two corpora counts once. Files and lines count the selected commits; each corpus indexes the whole repository at its pinned commit.
Selected work
gguatit/Anonymous_Chat
The corpus is the whole repository at the pinned commit, not only the files of the selected commit.
Not installable yet: this corpus is not in the hub registry until its licence review passes, so xerj corpus add and xerj ask cannot fetch it. It becomes installable when this card says live in the hub.
Sign in with GitHub or Hugging Face to see more: every selected commit in full (commit date, every folder, licence review state, pin) and the pack builds of each corpus.
[](https://hub.xerj.org/contributors/gguatit)
Included with the author's permission (email, 2026-10-14), under each repository's own licence, with attribution. Want it removed? Reply to the email or open an issue on xerj-org/xerj.