The package automatically synchronizes its agent skill into user-level agent directories during ordinary command startup. The content appears to be the CLI’s own skill, so this is an agent-extension lifecycle risk rather than evidence of credential theft or broader agent-setting hijacking.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
Running a normal CLI command after startup.
Impact
Agent behavior may change when those agents load the updated skill; this is an automatic agent-extension lifecycle risk.
Mechanism
Startup calls syncInstalledSkills(), which writes the package’s skill and references under the user’s Codex, Claude, Gemini, and Antigravity directories.
Evidence
index.jspackage.json~/.codex/skills/remits-cli/SKILL.md~/.codex/skills/remits-cli/references/*~/.claude/skills/remits-cli/SKILL.md~/.claude/skills/remits-cli/references/*~/.gemini/skills/remits-cli.md~/.gemini/skills/remits-cli/references/*~/.gemini/config/skills/remits-cli/SKILL.md