The inspected source shows a concrete agent-configuration write from authenticated service data during account-work commands, so a warning is appropriate. The package has no install lifecycle hook, and the inspected behavior does not establish a malicious payload or another concrete attack.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
Running a non-help, non-JSON account-work command such as components, test, token, tools, tool, or verify can trigger guide synchronization; guides sync can also be invoked directly.
Impact
Updated instructions may affect how AI agents behave when working in that repository.
Mechanism
The CLI fetches guide data from the configured Remits service using a saved session and writes its instruction field to repository-level agent instruction files.
Evidence
index.jsCLAUDE.mdAGENTS.mdGEMINI.md
Network endpoints1
remits-529558023549.us-east5.run.app