Skip to content

Roadmap

Ankit Parekh edited this page Nov 18, 2026 · 1 revision

This page is generated from ROADMAP.md. Edit the source in the repository and the change will be synced here.

Roadmap

This roadmap is a statement of intent, not a promise of dates. ReboundOps is maintained by one person, so items move when there is time or when a contributor picks them up. Each item comes from a limitation that is documented in docs/RELEASE_READINESS.md or in the release notes.

Now: making 0.2 dependable

Item Why Help wanted?
Windows 11 testing Windows is untested. File-permission hardening is best effort there. Yes: a test report, or a CI job
Intel macOS testing It is expected to work but has not been separately tested. Yes: run npm run verify:release and report
README screenshots The README has no screenshots yet. They must use fictional fixture data. Yes
Wider accessibility checks axe checks cover the main views and panels; the application detail page and a populated tracker are not covered yet. Yes
Community setup Discussions, the Wiki, a branch ruleset and Dependabot are prepared but not enabled. Maintainer

Next release

These items are planned in the release notes for the next release:

  • Better PDF layout handling without OCR, so multi-column resumes merge lines less often. Today a warning is shown.
  • More official sources where documented public APIs exist. Workday stays out of scope until it offers a documented public API.
  • An optional discovery schedule per provider, with the same consent, rate limits and caching as manual Discover.
  • An authorised export from the earlier version, so real data can move through the neutral import format. It will be done as a separate, isolated activity; see docs/LEGACY_EXPORT_PROCEDURE.md.

Later, or still being considered

  • Live verification of the cloud AI providers. They are tested with mocked requests only. Live checks need the user's own key and consent, so they cannot run in CI.
  • OCR for scanned PDFs, only if it can run locally, offline and at a reasonable size.
  • Cross-company search. None of the supported job boards offers it, so it would need a source with a documented API that permits it.
  • Pay comparison across currencies or periods. Today hourly pay is never converted, which avoids guessing.

Not planned

These would conflict with the project's principles:

  • automatically submitting applications, or marking them Applied without you;
  • sending emails or messages for you;
  • scraping behind logins, bypassing CAPTCHAs or anti-bot measures, or using undocumented private APIs;
  • an account system, telemetry or cloud sync run by the project;
  • AI features that write facts that are not in your records.

Proposing something

Open a feature request. Explain the problem first and the solution second, and say how it fits the principles above.

Clone this wiki locally