The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and design software based on user-provided inputs and code. This creates an attack surface where instructions embedded in the code being analyzed could influence the agent's behavior.
Ingestion points: Ingests code for analysis, design, and architecture tasks as described in SKILL.md and the localized description files.
Boundary markers: The skill does not define specific delimiters or instructions for the agent to ignore embedded prompts in processed code.
Capability inventory: While the skill itself is instructional, the agent's execution environment (such as Claude Code) typically possesses capabilities for file system operations, command execution, and network access.
Sanitization: There are no instructions for the agent to sanitize or validate external code before processing it.

software-architecture