The Agent Skills Directory

[SAFE]: The project initialization script (scripts/init-project.sh) performs only local file system operations, such as creating directories and copying templates. It uses standard utilities like sed for string replacement and does not make any network requests or attempt privilege escalation.
[SAFE]: The instructions in SKILL.md explicitly mandate a security protocol where the agent must filter sensitive information (such as API keys, tokens, and passwords) from configuration files before passing context to sub-agents.
[SAFE]: The skill utilizes the platform's native sessions_spawn mechanism for sub-agent orchestration in a transparent manner, consistent with its stated purpose of managing a multi-agent writing workflow.
[SAFE]: Analysis of all 75 files revealed no evidence of obfuscation, data exfiltration, or unauthorized persistence mechanisms. All project-related operations are scoped to the local workspace.

12agent-novel