swarm

SUSPICIOUS: the skill's core behavior matches its stated parallel-processing purpose, and install trust appears normal for official LangChain npm packages. However, its combination of large-scale untrusted file processing with optional full subagents that can edit/write files makes it a medium-risk agent skill, especially for prompt-injection-driven or overly broad automated code changes.