Skip to content
HomeNewsDigestsConceptsGuidesToolbox
AboutSubscribeUA
Subscribe

AI Today Brief

The daily AI-engineering brief. Built in public. EN · UA.

XTelegramLinkedInYouTubeRSS

Follow AI Today Brief on LinkedIn for daily AI-engineering updates and the weekly “5 shifts that changed how developers work” PDF.

Explore

NewsDigestsConceptsGuides

Company

SubscribeAdvertiseAbout

Legal

Editorial policyAI disclosurePrivacyTerms

© 2026 AI Today Brief. All rights reserved.

  1. Home/
  2. News/
  3. Tools & releases/
  4. Claude Code Issue Sweep Bot Auto-Closes 6,000 Reproducible Bug Reports
Tools & releases

Claude Code Issue Sweep Bot Auto-Closes 6,000 Reproducible Bug Reports

A misconfigured GitHub Actions workflow in the official Claude Code repository has auto-closed over 6,000 bug reports marked with has repro. The bot marks inactive issues as stale after 14 days and locks them permanently unless they reach ten upvotes.

September 15, 2026· 5 min read
OKCurated by Oleksandr Kuzmenko, AI Product Engineer·Updated September 15, 2026·Sources cited on every story
AI-assisted · editor-reviewed·How we use AI
Claude Code Issue Sweep Bot Auto-Closes 6,000 Reproducible Bug Reports

Impact: High

Why it matters

If you filed a Claude Code bug report, verify its status immediately and gather ten community upvotes before the automated workflow locks the discussion.

TL;DR

  • 01Over 6,000 Claude Code issues marked with has repro were auto-closed by bot workflows.
  • 02The stale bot requires 10 thumbs-up reactions (STALE_UPVOTE_THRESHOLD = 10) to exempt an issue from closure.
  • 03Closed issues lock after 7 days, preventing users from appending updated reproduction steps.

Key facts

Stale mark inactivity duration14 days
Thread lock delay after closing7 days
Auto-closed issues with reproduction steps
Over 6,000
Stale mark inactivity duration
14 days
Stale auto-close duration
14 additional days
Thread lock delay after closing
7 days
Upvote exemption threshold
10 thumbs up

Flawed Issue Lifecycle in Claude Code Repository

Issue #87647 reveals that the github-actions Issue Sweep workflow on anthropics/claude-code conflicts directly with public maintainer instructions. The official guidance stated that issues are closed only when missing reproduction data or a /feedback ID, and that commenting with an updated version allows re-opening without re-filing. However, the automated sweep triggers whenever human comments cease.

The 10-Upvote Threshold and Auto-Locking

Inspection of the repository actions reveals that STALE_UPVOTE_THRESHOLD is hardcoded to 10. Issues without at least 10 thumbs-up reactions encounter the following automated cascade:

1. Inactivity for 14 days triggers: nudge: "This issue has been automatically marked as stale due to inactivity." 2. Additional 14 days of silence triggers: nudge: "This issue has been marked for automatic closure." 3. The Lock Stale Issues workflow locks the thread permanently after 7 days of closure, making community follow-up impossible.

Because issues labeled with has repro or reproduced are not exempt from the 14-day timer, thousands of verified bug reports have been pruned silently.

Try it in 2 minutes

export const STALE_UPVOTE_THRESHOLD = 10;

typescript

✓ When to use

  • When filing bug reports, regressions, or CLI crashes on the official Claude Code repository.
  • When monitoring community pull requests and bug fixes in high-volume agent tooling repos.

✕ When NOT to use

  • When tracking bugs in personal or private repositories that do not use stale sweep actions.
  • When maintaining small open-source projects where every issue is triaged manually by humans.

What to do today

  • →Search your filed Claude Code issues on GitHub to confirm if they were marked stale or closed.
  • →Gather at least 10 thumbs-up reactions on active bug reports to hit STALE_UPVOTE_THRESHOLD.
  • →Include reproduction steps and Claude Version immediately in the initial issue description.
#Claude Code#GitHub Actions

Sources

  • [BUG] Over 6k issues labeled with has repro have been auto-closed since March 2026 #87647
ShareShare on XShare on LinkedIn
← Previous storyBenchmarking GPT-5.6 Luna vs GPT-6 Astra for Automated Code ReviewNext story →Capsule Packages Web Apps and Local Databases into Single Portable Files

Related stories

  • Tools & releasesBenchmarking GPT-5.6 Luna vs GPT-6 Astra for Automated Code Review
  • Tools & releasesCapsule Packages Web Apps and Local Databases into Single Portable Files
  • Tools & releasesRage4J Introduces Native LLM Response Testing for Java Stacks
  • Tools & releasesShopify Replaces React Native with Swift and Kotlin via Coding Agents

Email digest

Get the morning AI brief

One email a day — the stories that matter for engineers, founders and tech leads. Human-edited, with links to primary sources.

  • ✓120+ sources scanned daily
  • ✓Edited by a human
  • ✓1 email per day
  • ✓EN + UA

By subscribing you agree to the privacy policy.