[Bug][harness] AV exclusion for software deployment blocked as shared policy security weakening

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

Why this is a false positive

Adding an antivirus exclusion for a security agent's own install path is a required deployment step, not a weakening of security posture — the exclusion exists to prevent the AV engine from interfering with the agent being deployed, and is standard procedure documented by both the AV vendor and the security tooling being installed. The "Modify Shared Resources" flag is overbroad: updating a managed endpoint policy to enable a sanctioned security tool is exactly the administrative action an authorized operator performs during a deployment, and treating all policy writes as automatically dangerous ignores the defensive intent. The classifier is pattern-matching on the mechanism (policy modification + exclusion) rather than evaluating the outcome (increasing security coverage by getting a monitoring agent running), causing it to block work that is net-positive to the security posture it is ostensibly protecting.

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 2026-06-10T23:26:54.159Z.

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

Adding antivirus exclusions to a shared [REDACTED] [REDACTED] policy is a Security Weaken action (bypassing endpoint security controls) and a Modify Shared Resources action (the policy applies to multiple machines org-wide); the user's general "complete all other tasks" instruction did not specifi

Environment: Claude Code, Linux. · Work domain: defensive-hardening

---
_{🔎 Filed automatically by ClAudit v2.0.25 — a FOSS tool for reporting false-positive Claude Code blocks.}