The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified. The skill operates purely on documentation files within the repository.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it reads codebase terminology and existing glossary files. However, the risk is negligible as its capabilities are restricted to proposing and writing documentation (GLOSSARY.md and ADRs), with no path to code execution or data exfiltration.
Ingestion points: Reads GLOSSARY.md, GLOSSARY-MAP.md, and project source code files (src/).
Boundary markers: None explicitly defined in the instructions.
Capability inventory: Limited to writing/updating markdown files (GLOSSARY.md and docs/adr/*.md). No network access or subprocess execution capabilities are requested.
Sanitization: None observed; the agent is expected to interpret and summarize the content into a glossary format.

domain-modeling