The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied build metadata such as build names, stack descriptions, and key learnings, and interpolates them directly into multiple artifact templates including MDX logs, LinkedIn posts, and video demo scripts.
- Ingestion points: User-provided inputs defined in the inputs block of SKILL.md.
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the processed inputs.
- Capability inventory: The skill performs file system writes across multiple directories (content/, drafts/, skills/), executes a bash grep command for linting, and suggests using npx for rendering diagrams.
- Sanitization: No explicit sanitization or validation logic is applied to the input values before they are interpolated into the final artifacts.
[REMOTE_CODE_EXECUTION]: The skill documentation in resources/diagram-prompt.md provides an execution pattern using npx to download and run the @mermaid-js/mermaid-cli package.
[EXTERNAL_DOWNLOADS]: The skill fetches the Mermaid CLI tool from the official npm registry during the diagram rendering phase.
[COMMAND_EXECUTION]: The skill executes a bash grep command to perform automated linting of the generated files to ensure compliance with the voice specification defined in resources/voice-spec.md.
[DYNAMIC_EXECUTION]: The skill utilizes npx for runtime package execution to convert Mermaid source files into SVG diagrams.

Agentic Builder Lab