The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill attempts to read the user's shell configuration files, which are known to store sensitive credentials and environment variables.
Evidence: grep "alias saas-factory" ~/.zshrc and grep "alias saas-factory" ~/.bashrc in SKILL.md.
[COMMAND_EXECUTION]: The skill executes potentially destructive file system operations using the Bash tool on paths determined by the contents of user configuration files.
Evidence: rm -rf .claude/ and cp -r [RUTA_REPO_SF]/saas-factory/.claude/ .claude/ in SKILL.md.
[REMOTE_CODE_EXECUTION]: The skill uses Git to pull code from a remote source into the local environment without verifying the source's identity or content integrity.
Evidence: git pull origin main in SKILL.md.

update-sf