[Bug][harness] ClAudit: auto-mode classifier denied — The action reads the MSAL token cache to extract a user refresh token

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

Why this is a false positive

The block fired on a standard authentication flow: the user was explicitly walked through and completed an interactive device-code login, and the immediately following action only reads the token cache that the official CLI's own login wrote and uses it under that CLI's well-known public client ID — the documented, supported way that tool authenticates. No credential was exfiltrated, escalated, or used outside the scope the user just consented to; reusing a cached refresh token under the tool's own app ID is the normal mechanism, not a workaround. The "without explicit user authorization" rationale is incorrect here, since the user directly approved the login and issued "done, go" to proceed.

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 2026-06-13T06:25:35.844Z.

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

The action reads the MSAL token cache to extract a user refresh token and re-uses it under the [REDACTED] CLI's app ID — credential exploration beyond the agent's normal credential scope, without explicit user authorization for this specific approach.

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

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