The Agent Skills Directory

[SAFE]: The skill contains technical documentation and YAML/Markdown templates for agent orchestration. No malicious patterns such as credential theft or unauthorized network access were found.\n- [INDIRECT_PROMPT_INJECTION]: The skill architecture facilitates communication between agents via a message inbox system and shared tasks. This creates a surface for indirect prompt injection if teammates process malicious content. Agents have access to capabilities like Bash and Write. The documentation suggests 'Plan Approval' workflows to mitigate these risks.\n- [COMMAND_EXECUTION]: The hook configurations in assets/hook-examples.yml allow for executing shell scripts like npm test during task transitions. These examples use local project paths and standard development tools.

agent-teams