The Agent Skills Directory

[SAFE]: The skill focuses on providing architectural advice and decision trees for optimizing SaaS payloads and database storage. All external references are to well-known standards organizations (IETF, RFC Editor) or official documentation (Mozilla MDN, PostgreSQL), which are trusted sources.
[INDIRECT_PROMPT_INJECTION]: The skill describes processes for summarizing external data (logs, code, API responses) to reduce token count. While processing untrusted data is a potential attack surface for indirect prompt injection, the skill provides logical constraints (preserving intent, outcomes, and evidence paths) and does not instruct the agent to execute or evaluate the summarized content in a dangerous manner. The capability is restricted to standard agent operations.

compression