lore-setup
Warn
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: MEDIUMPERSISTENCEPRIVILEGE_ESCALATION
Full Analysis
- [PERSISTENCE]: The skill instructs the agent to write rules into global configuration files, such as
~/.claude/CLAUDE.md,~/.codex/AGENTS.md, and~/.qwen/QWEN.md. This ensures that the injected instructions remain active and influence every future session of the AI agent on the host system. - [PRIVILEGE_ESCALATION]: The skill targets files outside the immediate project workspace by directing the agent to modify hidden configuration files within the user's home directory. This behavior expands the skill's control from a local project context to the entire environment of the agent CLI.
Audit Metadata