The Agent Skills Directory

[NO_CODE]: The skill is comprised entirely of text-based documentation and templates. There are no scripts, binaries, or tool configurations that could lead to code execution or system modification.
[EXTERNAL_DOWNLOADS]: The skill contains links to public LinkedIn posts for methodological reference. These links point to a well-known service and are provided for context; the skill does not perform automated fetching or execution of content from these URLs.
[INDIRECT_PROMPT_INJECTION]: The skill generates drafts based on user-provided business details, creating an ingestion surface.
Ingestion points: User-provided data enters via the 'commercial brief' in SKILL.md and is formatted using templates/output-template.md.
Boundary markers: The templates utilize markdown headers and bracketed placeholders to distinguish between model instructions and user-supplied data.
Capability inventory: No capabilities for network access, file writing, or command execution are defined within the skill.
Sanitization: The skill relies on the agent's internal text processing and does not specify additional sanitization, which is consistent with a text-only generation task.

koval-linkedin-authority