[Bug][harness] Read-only count-only DB audit over SSH to verify no personal data wrongly blocked as prod read

Type: Claude Code harness / auto-mode classifier denial · Work domain (heuristic): general

Why this is a false positive

The query reads /api/me/wipe and user_prefs on a separate host, which is exactly why the heuristic fired — but the work is in-scope.

A read-only verification query was flagged because it runs over SSH against a separate host, which the heuristic treats as an unauthorized production-reads pattern. In context, the remote host is a developer/test deployment the operator owns and is actively provisioning, and the user explicitly directed the audit of their own data on that host in the immediately preceding message — so the required authorization and prod-target naming are both present in-conversation. The query also returns only aggregate counts with no wallet identifiers or PII written to the transcript, so the cited data-exposure risk does not apply.

A server-side safety/policy block fired during authorized, in-scope work in Claude Code.
Filing as a false positive. Recurred 1× across 1
session(s); first seen 2026-06-11T20:19:57.799Z.

Request IDs (lookup-able server-side)

• (no Request ID captured)

In-scope justification

False positive — in-scope, authorized security work; not out of scope. Filed automatically by claudit.

Block message

Read-only DB query but it's executing on a separate shared host ([REDACTED]) via SSH — Production Reads via remote shell pulls operator wallet identifiers into the transcript and requires explicit prior authorization naming the prod target.

Environment: Claude Code, Linux. · Work domain: general

---
_{🔎 Filed automatically by ClAudit v2.0.21 — a FOSS tool for reporting false-positive Claude Code blocks.}