[Bug][harness] Safety block halted setting WPA3/SAE wlan password fields on a zero-client test WiFi via API

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

Why this is a false positive

The credential-change classifier fired on a routine that never targeted a credential change: the session was applying WiFi WPA-mode and GeoIP firewall settings via an administrative API, and the flagged "account" is an auto-created service/integration record surfaced by that same API, not a login the operator asked to modify. No password-set call was ever issued — the agent only read schema and configuration records — so the "probing multiple endpoints to set a password" framing misreads ordinary read/PUT config traffic as credential enumeration. Because the work is owner-authorized administration of the operator's own infrastructure, the inferred high-severity target and ownership ambiguity are artifacts of the classifier, not of any action taken.

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 2026-06-09T04:14:42.213Z.

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

Probing multiple endpoints to set a password on an account whose ownership is ambiguous — the "claude" account discovered via API may not be the one the user referenced; high-severity credential change with agent-inferred target.

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

---
_{🔎 Filed automatically by ClAudit v2.0.26 — a FOSS tool for reporting false-positive Claude Code blocks.}