The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches configuration files from arbitrary GitHub repositories identified through a third-party search service. While the domain raw.githubusercontent.com is a standard service, the specific repository content is user-controlled and not subject to verification.
[INDIRECT_PROMPT_INJECTION]: The skill has a significant attack surface for indirect prompt injection due to the ingestion and analysis of untrusted third-party data.
Ingestion points: Untrusted data enters the agent context via WebFetch calls to kubesearch.dev and raw.githubusercontent.com as described in the Workflow section of SKILL.md.
Boundary markers: Absent. There are no instructions to use delimiters or to ignore potential instructions embedded within the fetched values.yaml files.
Capability inventory: The skill uses WebFetch for network operations to retrieve data from external sources.
Sanitization: Absent. The skill does not describe any validation or filtering mechanisms for the external content before it is processed by the agent.

kubesearch