Git diff chip counts lines of untracked files marked -diff in .gitattributes #16611

Activity

1. Ga-3tan

   added

   bugSomething isn't working.Something isn't working.

   on Oct 22, 2026
2. warp-for-oss commented on Oct 22, 2026

   

   warp-for-ossBot

   on Oct 22, 2026

   Last edited by warp-for-oss

   Contributor

   More actions

   @Ga-3tan

   Here's what I found while triaging this issue:

   The report has enough detail to reproduce, and the behavior matches Warp's current Git diff stats code path. The expected fix is for the prompt/top-right diff stats to respect Git attributes for untracked files the same way the Code Review panel does.

   I've flagged this for the factory auto-implement queue.

   Maintainer details

   I concluded that untracked files marked -diff or binary in .gitattributes are counted by Warp's git diff stats metadata even though the Code Review diff path treats them as binary, causing the prompt/top-right diff counts to disagree with the Code Review panel..

   Triage summary

   This looks like a real bug in the Git diff stats shown outside the Code Review panel. The reporter's explanation matches the current code path: untracked files are enumerated separately from git diff --numstat HEAD, and when there is no numstat row for an untracked file Warp falls back to directly counting working-tree lines.

   Reproducibility

   High. The issue includes minimal reproduction steps, screenshots, and environment details. A local Git check with an untracked generated.txt marked -diff shows that git diff --numstat HEAD produces no entry, while git diff --no-index /dev/null generated.txt treats the file as binary because the diff attribute is unset.

   Likely root cause

   app/src/code_review/diff_state/local.rs computes aggregate diff metadata in diff_metadata_against_head and diff_metadata_against_specific_branch. For untracked files missing from git diff --numstat, it calls num_lines_in_file_if_non_binary, which only inspects file bytes and size and does not consult .gitattributes. app/src/code_review/git_repo_model/local.rs feeds those aggregate stats into GitStatusMetadata, and app/src/context_chips/current_prompt.rs / app/src/context_chips/display_chip.rs render that metadata in the prompt diff chip. The Code Review diff view uses a separate per-file path that runs git diff --no-index for untracked files and therefore treats the -diff file as binary, so the displayed totals can diverge.

   Duplicate check

   I found related open reports about stale or inconsistent Git diff chips, including #12212, #15209, #16319, #9787, #12664, and #12674, but none of the scanned open issues or recently closed issues matched the specific .gitattributes -diff / untracked-file line-count scenario closely enough to mark this as a duplicate under the two-candidate duplicate threshold.

   Factory auto-implementation

   Eligible. The behavior is a client-side bug with high reproducibility, a known root cause, an unambiguous definition of done, and a localized fix path in Git diff metadata handling.

   This is my automated analysis and may be incorrect. A maintainer will verify the details.

   Powered by Oz
3. warp-for-oss

   added

   agent:bugBugs/regressions. Assigned by Warp triage agent.Bugs/regressions. Assigned by Warp triage agent.

   area:code-reviewGit diff views, review UI, review comments, and PR-focused agent flows.Git diff views, review UI, review comments, and PR-focused agent flows.

   area:shell-terminalTerminal input/output, shell integration, prompt behavior, and block rendering.Terminal input/output, shell integration, prompt behavior, and block rendering.

   factory-auto-implementQueued for automated implementation by the Warp factoryQueued for automated implementation by the Warp factory

   os:macmacOS-specific behavior, regressions, or requests.macOS-specific behavior, regressions, or requests.

   repro:highThe report includes enough evidence that the issue appears highly reproducible.The report includes enough evidence that the issue appears highly reproducible.

   triagedIssue has received an initial automated triage pass.Issue has received an initial automated triage pass.

   on Oct 22, 2026
4. warp-agent-staging

   assigned

   kevinyang372

   on Oct 22, 2026
5. warp-agent-staging

   added

   warp:auto-triage-review

   on Oct 22, 2026