The Agent Skills Directory

[SAFE]: The skill is entirely instructional and provides strategic guidance on context engineering for AI-assisted development. It does not contain executable code, scripts, or network operations.
[INDIRECT_PROMPT_INJECTION]: The skill explicitly addresses the threat of indirect prompt injection by advising the agent to treat instruction-like content from untrusted external sources (documentation, third-party API responses) as data to be surfaced to the user rather than directives to be followed.
[DATA_EXPOSURE]: The provided configuration templates include security best practices, such as explicit instructions to never commit .env files or secrets to source control.
[NO_CODE]: The skill does not bundle any scripts, binaries, or automated tasks, relying instead on high-level guidance and example templates for standard project configuration files.

context-engineering