The Agent Skills Directory

[COMMAND_EXECUTION]: The skill creates and executes shell scripts (bin/conductor-setup and script/server). It uses chmod +x to grant execution permissions and ln -sf to create symbolic links to sensitive files including .env and config/master.key from the path provided in the environment variable CONDUCTOR_ROOT_PATH.
[EXTERNAL_DOWNLOADS]: The setup script runs bundle install and npm install, which download third-party code libraries from the RubyGems and NPM registries respectively.
[PROMPT_INJECTION]: The skill reads and modifies existing Ruby and YAML configuration files. This presents an indirect prompt injection surface where maliciously crafted comments or code in those files could attempt to influence the agent's behavior during the update process.
Ingestion points: config/initializers/sidekiq.rb, config/cable.yml, config/environments/development.rb, config/initializers/rack_attack.rb.
Boundary markers: Absent.
Capability inventory: Shell command execution via Bash tool, file creation, and file modification.
Sanitization: No explicit sanitization or validation of the ingested file content is performed before processing.

conductor-setup