The Agent Skills Directory

[PROMPT_INJECTION]: The skill cross-references project source code and documentation to maintain a domain model, which creates an ingestion surface for indirect prompt injection. Malicious instructions hidden in the codebase could attempt to influence the agent's behavior during analysis. Ingestion points: Processes files within the src/ directory and documentation files like CONTEXT.md or ADRs. Boundary markers: No specific delimiters or safety warnings are defined to isolate the ingested code content from the agent's instructions. Capability inventory: The skill's primary capabilities are reading from the filesystem and writing to documentation files; it does not request network access or execute arbitrary system commands. Sanitization: The instructions do not specify any validation or sanitization of content read from the project files.

domain-modeling