The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @cuelight/cli tool from the NPM registry using the command bun add -g @cuelight/cli. This package is the primary interface for the skill's production workflow.
[COMMAND_EXECUTION]: The skill's functionality is centered around executing shell commands via the cuelight-cli. These commands handle data management, asset generation (images and videos), and project state tracking.
[PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection (Category 8) as it ingests and processes untrusted user data such as novel scripts or source documents.
Ingestion points: User-provided files are loaded into the workflow via commands like cuelight-cli source draft create-from-file and cuelight-cli episode set-script.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the prompt templates when interpolating content.
Capability inventory: The agent has access to shell execution, local file read/write within the ./.cuelight/ directory, and network operations via the CLI tool.
Sanitization: Content sanitization is managed by the external CLI tool and the backend service it connects to.

cuelight-drama