The Agent Skills Directory

[PROMPT_INJECTION]: The skill analyzes external code and content to generate explanations. This ingestion process lacks explicit boundary markers or instructions to disregard potential commands within the target data. Consequently, malicious instructions hidden in code comments or documentation could influence the agent's behavior during analysis.
Ingestion points: Target code is accessed via file-reading tools and the Explore subagent as defined in references/explain.md.
Boundary markers: Prompts for subagents do not use delimiters (such as XML tags or triple quotes) to isolate and treat the target content as data rather than instructions.
Capability inventory: The skill possesses capabilities for file system access (Read, Grep, Glob), file modification (Write), and subagent delegation (Task tool).
Sanitization: No input validation or filtering is specified for the code or documentation being analyzed.

code-explanation