Source supports explicit user-command agent configuration capability, which warrants a warning under the supplied policy. No unconsented install-time agent mutation, command-output exfiltration, or other concrete malicious chain was established.
Static reason
One or more suspicious static signals were detected.
Trigger
Running toolkit init or toolkit skills install; a generated prepare hook can subsequently refresh project skills.
Impact
Changes instructions available to AI agents; linked skills follow package updates and project copies can refresh through prepare.
Mechanism
Appends a conventions pointer, creates skill files, and installs links or copies with saved installation state.
Evidence
dist/cli.jsAGENTS.mdpackage.json.agents/skills.claude/skills.cursor/skills.toolkit-skills.json
Network endpoints1
registry.npmjs.org/@fxri%2Ftoolkit/latest