The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill incorporates a workflow to fetch external library information using the Context7 documentation tool.
Ingestion points: Output from the context7_query-docs tool is ingested into the agent's context as documented in SKILL.md.
Boundary markers: There are no specific delimiters or instructions to ignore embedded commands within the retrieved documentation content.
Capability inventory: The skill provides instructions for project construction, file writing, and shell command execution (mvn spring-boot:run, mvn test), which could be targets for instructions embedded in poisoned documentation.
Sanitization: No sanitization or validation of the retrieved external content is performed before it is added to the prompt.
[SAFE]: The skill explicitly teaches secure coding practices, specifically warning against hardcoding credentials in references/configuration.md and demonstrating the use of BCryptPasswordEncoder in references/security.md.
[EXTERNAL_DOWNLOADS]: External references point exclusively to official Spring Framework resources (spring.io), which are well-known and trusted service domains.

java-springboot