Skip to content

router: follow-ups for the store, forms and router inspectors #368

Closed

Description

@erkamyaman

Items still open from the audit issues #347, #348 and #349. The fixed items landed in #362; this issue tracks what is left. Each item says why it was not part of that fix.

From #347 (NgRx store)

Partly done (the rest is open):

  • Low: Classic state is serialized twice; same() compares with JSON.stringify: report() now reuses the serialized classic state (c.last) while the store still holds the same object as lastRaw. same() is now a structural deep compare with no string building (and key order no longer matters). Not done: the overlay's JSON.stringify body compare in ngrx-overlay.ts, because removing it needs dirty tracking for signal stores too. Caveat: a store that mutates its state in place would now report the cached copy, which was already invisible to the diff. (done in fix(overlay): keep console call sites, gate the Native profiler, skip unchanged NgRx stores #466)
  • Low: Report entries are not shape-checked (null log entry throws in mergeNgrxReport): mergeNgrxReport now checks each store, log entry and classic block with valibot looseObject schemas and drops bad ones (a null entry no longer throws). Test added. Not done: the per-page size cap, because honest pages already bound blobs through serialize budgets, and choosing what to show for an oversized blob is a product decision. (done in feat: add cdSlowMs, an NgRx size cap, transfer cache options, security status and token rotation #471)

Open:

From #348 (forms and WebMCP)

Partly done (the rest is open):

Open:

From #349 (router)

Partly done (the rest is open):

Open:

Tracked elsewhere

These items from the audit are not listed above because another issue already tracks them:

Activity

  1. added
    bugSomething is broken or shows wrong data
    featureA feature request or a pull request that adds one
    on Oct 12, 2026
  2. erkamyaman commented on Nov 7, 2026

    @erkamyaman
    MemberAuthor

    The forms items landed in 1.0.0 with #479 (squash commit 2baa859). The commit message says "Refs #367" by mistake; it belongs to this tracker. Minor findings from its review are collected for the pre-release cleanup PR.


    Generated by Claude Code

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

    bugSomething is broken or shows wrong datafeatureA feature request or a pull request that adds one

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions