The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface where untrusted data from external sources could potentially override agent instructions.
Ingestion points: The skill retrieves post titles, metadata, and body content from Naver Blog via list_naver_drafts and update_post. It also explicitly instructs the agent to relay userMessage content from the bridge_status tool and various error responses verbatim to the user context.
Boundary markers: The instructions do not prescribe the use of delimiters (such as XML tags or triple backticks) or specific directives for the agent to ignore instructions that might be embedded within the retrieved data.
Capability inventory: The agent possesses the capability to modify external state by publishing drafts or rewriting live blog posts through the publish_naver_draft and update_post tools.
Sanitization: There is no requirement or logic defined to sanitize, filter, or validate the external content or bridge messages before they are processed or presented by the agent.

naver-drafts