[Bug][harness] ClAudit: auto-mode classifier denied — Running build-agent.sh extracts codesign/relay signing credentials and

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

Why this is a false positive

I'll write that text for you.

The block was triggered on an authorized fleet-provisioning task: build-agent.sh is the project's normal, in-repo build wrapper that compiles a monitoring/management agent and signs it with the operator's own credentials for deployment to machines the operator controls. The classifier misread a routine signed-build-and-deploy step as malicious credential exfiltration and unauthorized agent "arming," when in context it is standard first-party CI/build tooling run by the infrastructure's owner. The user's prior frustrated remark was directed at the false positive itself, not an attempt to authorize abuse — no evasion of any control was performed or required.

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 2026-06-24T03:04:26.668Z.

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

Running build-agent.sh extracts codesign/relay signing credentials and mass-builds agent binaries for fleet-wide deployment across all client machines — credential extraction into build artifacts plus production/shared-infra agent arming, pushed through after the user's explicit "classifier is a cun

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

---
_{🔎 Filed automatically by ClAudit v2.0.7 — a FOSS tool for reporting false-positive Claude Code blocks.}