The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes externally provided or recently modified code snippets, creating a surface for potential indirect prompt injection attacks. \n- Ingestion points: The agent is instructed to analyze and optimize recently modified code within the session (specified in SKILL.md workflow). \n- Boundary markers: The instructions do not specify the use of delimiters (such as XML tags or markdown code blocks) or explicit 'ignore embedded instructions' warnings for the code being processed. \n- Capability inventory: The skill is limited to providing code recommendations and text output; it does not request access to shell execution, network operations, or file-writing tools. \n- Sanitization: No input validation or content filtering of the code snippets is implemented to detect or strip instructions hidden in comments or strings.

code-simplifier