The Agent Skills Directory

[SAFE]: The skill performs legitimate configuration management tasks using provided tools from the author's infrastructure.
[PROMPT_INJECTION]: Identifies a surface for potential indirect prompt injection as the skill manages external prompt content. This is a functional requirement of the skill's purpose.
Ingestion points: Reads snippet text and configurations via get-prompt-snippet and get-ai-config (SKILL.md).
Boundary markers: No explicit prompt delimiters are used to wrap snippet content during verification or listing.
Capability inventory: Includes the ability to write and update snippets and configurations (update-prompt-snippet, update-ai-config-variation).
Sanitization: No sanitization of ingested snippet content is specified in the instructions.

aiconfig-snippets