The Agent Skills Directory

[SAFE]: The skill is composed of purely instructional documentation and YAML metadata. It contains no executable scripts, automated commands, or network-active tools.
[INDIRECT_PROMPT_INJECTION]: The instructions suggest reading local project documentation such as GLOSSARY.md and Architecture Decision Records (ADRs) to ensure consistent terminology. While this constitutes an ingestion of external data, it is a standard practice for development agents and does not present an exploitable risk in this context.
[CREDENTIALS_UNSAFE]: Code examples correctly demonstrate the use of environment variables (e.g., process.env.STRIPE_KEY) for secret management, which is a recommended security best practice.

tdd