The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from the codebase and provides architectural guidance based on that content.
Ingestion points: The skill reads files from the src/ directory and existing documentation such as GLOSSARY.md and architecture decision records to cross-reference domain terminology and architectural decisions.
Boundary markers: No explicit delimiters or boundary markers are defined for isolating untrusted codebase content within the agent's context.
Capability inventory: The skill has file-writing capabilities to create and update GLOSSARY.md, architecture decision records (ADRs), and GUIDELINES.md.
Sanitization: There is no evidence of specific sanitization or filtering applied to the ingested content before it is processed or written back to the filesystem.
[SAFE]: The skill's behavior is consistent with its stated purpose of managing project documentation and domain models. The analyzed instructions and templates contain no evidence of command execution, unauthorized network exfiltration, credential harvesting, or obfuscation techniques.

domain-modeling