The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscated commands, or security vulnerabilities were detected in the skill instructions or reference documents. The skill is purely instructional and used for creative text transformation.
[NO_CODE]: The skill is composed entirely of Markdown documentation and does not execute scripts, install software, or depend on external packages.
[SAFE]: The skill handles untrusted user input ("raw ideas") but lacks the system-level capabilities required for an Indirect Prompt Injection attack to pose a threat. Mandatory evidence chain: 1) Ingestion points: "raw ideas" and "input concepts" (SKILL.md); 2) Boundary markers: uses structural XML tags to delimit scenes; 3) Capability inventory: no network, file system, or code execution capabilities detected; 4) Sanitization: includes specific guidance on XML special character escaping (references/pipeline-integration.md).

screenwriter