The Agent Skills Directory

[COMMAND_EXECUTION]: The skill creates and executes shell scripts (bin/conductor-setup, script/server) and uses chmod +x to set permissions. It also runs bundle install and npm ci to manage project dependencies. The instructions emphasize using --ignore-scripts for npm to mitigate risks from lifecycle scripts.
[DATA_EXPOSURE]: The setup script facilitates the symlinking of sensitive development files, specifically .env and config/master.key, into the workspace. The skill includes explicit warnings that these secrets should be handled carefully, kept out of version control, and used only in approved development contexts.
[EXTERNAL_DOWNLOADS]: The skill initiates network requests via bundle install and npm ci to download third-party packages. It advises the user to review lockfiles (Gemfile.lock, package-lock.json) before proceeding.
[INDIRECT_PROMPT_INJECTION]: The skill reads existing Rails configuration files (Ruby and YAML) and modifies their content to interpolate REDIS_URL environment variables. While this processes external file content, the modifications are constrained to specific Redis configuration blocks.

conductor-setup