[Bug][harness] Safety block halted authorized SSH redeploy of forensics service after binary rebuild

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

Why this is a false positive

The block fired on a routine application deployment — building a binary and restarting its backend service over an established SSH alias — which the user had already directed across the immediately preceding turns as the explicit next step of an active development task. The "production deploy without authorization" framing misreads in-scope, user-initiated iterative work as an unsanctioned outward-facing action; the deploy target, transport, and intent were all set up earlier in the same session at the user's request. Restarting a service the user is actively developing is a normal step in the build-deploy-verify loop, not an unauthorized production change.

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 2026-06-11T19:53:23.420Z.

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

Restarting [REDACTED] on the shared production CT ([REDACTED]) is a production deploy action via remote shell without explicit user authorization for this specific deploy in the current turn.

Environment: Claude Code, Linux. · Work domain: malware-forensics

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