bmad-story-automator

This is agent-directed operational text rather than executable code. It contains prompt-injection-like instructions encouraging autonomous continuation and state modification, which should be treated as untrusted and not followed without authorization. No direct malware behavior is demonstrated.

SUSPICIOUS. The visible skill is too incomplete to verify, with its real behavior deferred to an unseen workflow file. External evidence suggests legitimate BMAD-related tooling, but the combination of hidden instructions, packaged binary execution, unclear publisher alignment, and guidance to run with dangerously skipped permissions makes the footprint higher risk than the minimal skill text suggests.