Skip to content

Findings: bulk status update, and a report stage separate from the triage status #297

Closed

Description

@BaaRai

Problem

Triaging a batch of findings means opening each one and changing its status one at a time. With ten or twenty findings from a single task that is several minutes of clicking, and it is easy to miss one.

The same workflow also needs a stage after triage: "this report has been submitted to the platform, waiting for review". Today that state has nowhere to live.

Proposal

  1. Bulk status update on the Findings page. Select rows with the existing checkboxes, pick a status, apply. Server side: one endpoint, one transaction, one result per id. This lands first.
  2. A report stage, separate from the triage status. A report_state field with a fixed set (none | submitted | accepted | rejected | duplicate), editable in bulk from the same bar, so internal statistics can count submitted and accepted findings without losing the verdict.

Why not a submitted status

status is the triage verdict. The RSI lesson extraction reads it: confirmed/resolved/fixed add positive evidence, false_positive/duplicate/ignored/risk_accepted add negative evidence, and a reversal retracts the other kind. A submitted value would fall between those branches. It would also replace confirmed in the same column, so a finding could no longer be both confirmed and submitted, and the statistics would count a workflow stage as a verdict.

A separate field keeps both facts, leaves the lesson logic untouched, and gives clean numbers: how many are confirmed, how many of those are submitted, how many are accepted.

Not doing: user-defined status sets

Status values are enumerated in Go (db.ValidFindingStatus), in the TypeScript union, in labels and tones, in report filters and in the lesson switch. Making them user-editable means a status table, references, dynamic labels, and rewriting all of those points, for little benefit in a single-admin tool. If free-form classification is the need, tags are the safer direction.

Notes for the bulk endpoint

  • Cap the batch, dedupe ids, and validate the status with the same validator as the single-row endpoint.
  • One transaction for the batch, with a per-id result so a missing id does not hide the rest.
  • One notification event per changed finding, same semantics as the single row. A digest can come later if it is noisy.
  • Reuse the node payload mirror added in fix(db): mirror finding triage onto the exploration node #289, so the per-task view stays in sync with the findings table.

Activity

  1. BaaRai commented on Nov 17, 2026

    @BaaRai
    OwnerAuthor

    Both halves are in main: the bulk status update in d4c1a5c (#298), and the report stage as a separate report_state field in 9a2a9e6 (#308). "submitted" stayed out of the triage status on purpose: the fixed set none/submitted/accepted/rejected/duplicate is bulk-editable and readable in the list, the filters, the stats and the export. Closing.

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

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions