Skip to content

Learn/suggest words from user's own typing history (personal vocabulary) #817

Description

@Rahulsharma0810

Problem

Cotabby's suggestions are based on the live caret prefix plus static, manually-entered settings (name, languages, custom writing-style rules — see WelcomePersonalizeStepView, SuggestionSettingsModel). There is no mechanism that learns from what the user has actually typed/accepted over time, so it never starts suggesting a user's own recurring words, names, jargon, or phrasing back to them.

Request

Add a privacy-first, local, opt-in mechanism that learns from the user's own typing/accepted-completion history and feeds that back into suggestions (e.g. as part of "Use My Instructions" / custom-rules context), so words the user has already typed repeatedly get suggested proactively.

Related open issues (please upvote/subscribe if this matters to you)

Filing this as a distinct, findable issue that ties the above together, since the underlying capability (personal vocabulary/word-history learning) still doesn't exist in main as of this writing (verified via source: no PersonalizationStore/learnedPreference/typingHistory anywhere in the codebase).

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    • Status
      Backlog

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions