The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified. The skill implements productivity tools using local scripts and established copywriting frameworks.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data such as user drafts and external URLs as part of its core functionality. It mitigates risks by including specific security-conscious instructions for URL ingestion and providing automated linting for structured metadata. * Ingestion points: User-provided drafts (humanize.md) and external URLs (video.md, podcast.md). * Boundary markers: The skill enforces editorial integrity through verification checklists ensuring facts and voice are preserved. * Capability inventory: Access to bash, run_command, and file manipulation tools. * Sanitization: Implements frontmatter-guard.ts for YAML syntax validation and aeo-schema.ts for structured data generation.
[COMMAND_EXECUTION]: The skill uses run_command and bash to execute local TypeScript scripts via the bun runtime. These scripts are part of the skill's distribution and perform administrative tasks like metadata linting and schema generation.

content