Public research record · Editorial guide updated 2026-10-29

Claude subagent's test deleted 34 sibling project directories

Review RAT-085 · Catalogue RAT-085

RogueWatchAwaiting evidence

Spicker’s revised first-person account reconstructs an October 5 guard relaxation, destructive unit test, and deletion across 34 project directories. Most work was recovered; unpushed history was not all recovered. No model formed an independent deletion objective. Real unintended deletion remains a lead, but organizational victim and second-domain substantive corroboration are unestablished.

Reviewed: · Rule: roguewatch-organizational-harm-v1

Dates and discovery

Incident date
Not disclosed
Added to research / record
Posted to RogueWatch
No linked public incident
First observed
2026-10-30T13:13:36.979942Z UTC
Discovery channel
User-supplied comparison audit
Discovery source
https://rogueaitracker.com/incidents/claude-subagent-deleted-sibling-checkouts/

When to reconsider

New first-party traces or substantive corroboration from another publisher clarifies the identified authorization, organizational-victim, chronology, or outcome gap.

Dates use America/Los_Angeles. For older entries, Posted to RogueWatch uses the saved record addition date; a separate deployment time was not recorded. Future posting timestamps record first public qualification. Source publication and incident occurrence are separate.

Sources

Decision history

  1. · Case-by-case comparison audit

    RAT-085: awaiting-evidence. Spicker’s revised first-person account reconstructs an October 5 guard relaxation, destructive unit test, and deletion across 34 project directories. Most work was recovered; unpushed history was not all recovered. No model formed an independent deletion objective. Real unintended deletion remains a lead, but organizational victim and second-domain substantive corroboration are unestablished. Source checks: read: https://sebastianspicker.github.io/posts/the-day-claude-deleted-my-codebase/. Spicker’s revised first-person account reconstructs an October 5 guard relaxation, destructive unit test, and deletion across 34 project directories. Most work was recovered; unpushed history was not all recovered. No model formed an independent deletion objective. Real unintended deletion remains a lead, but organizational victim and second-domain substantive corroboration are unestablished.

    Evidence source
  2. · Retained review-queue assessment

    Spicker’s revised first-person account reconstructs an October 5 guard relaxation, destructive unit test, and deletion across 34 project directories. Most work was recovered; unpushed history was not all recovered. No model formed an independent deletion objective. Real unintended deletion remains a lead, but organizational victim and second-domain substantive corroboration are unestablished.

    Evidence source

A review decision applies only to the named scope. Confidence in a report and qualification under the inclusion rules are separate questions. Review records do not add incidents to the globe or leaderboard.