The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions are focused on assisting the user with backend tasks. No attempts to override agent safety guidelines, extract system prompts, or bypass restrictions were detected.
[DATA_EXPOSURE]: There are no hardcoded credentials or sensitive file path access. The skill explicitly instructs the agent to use environment variables for secrets and avoid hardcoding sensitive information like JWT secrets or API keys.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns (such as piping curl output to a shell) were identified. The external links provided are for documentation purposes and target well-known, trusted technology domains (e.g., nestjs.com, redis.io, postgresql.org).
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it ingests user-provided requirements and code to generate or optimize backend systems. However, this is inherent to its primary purpose as a development assistant. It mitigates risk by instructing the agent to follow strict security constraints (like input validation and parameterized queries) in the code it produces.
[OBFUSCATION]: No obfuscated code, Base64-encoded strings representing commands, or hidden characters were found.
[DYNAMIC_CONTEXT_INJECTION]: No use of shell command execution placeholders (e.g., backtick commands) was found in the SKILL.md file.

backend-development