The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and process untrusted user data from work sessions or research.
Ingestion points: Data enters the agent's context through user-provided session notes, research findings, and artifacts referenced in SKILL.md.
Boundary markers: The skill uses standard Markdown headers but lacks explicit safety delimiters or instructions for the AI to disregard potential commands hidden within the ingested content.
Capability inventory: The skill includes instructions to save documents to the local file system (Step 5 in SKILL.md).
Sanitization: There is no explicit logic provided for sanitizing or validating user-provided content before it is incorporated into the final briefing documents.
[COMMAND_EXECUTION]: The skill facilitates file system operations by instructing the agent to save generated briefings to the local directory structure (docs/briefings/).
[EXTERNAL_DOWNLOADS]: The file references/format-research.md contains informational links to well-known business and technology resources, including Atlassian, Amazon, and HashiCorp. These are recognized as safe, reputable sources for document templates and research.

create-briefing