The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches query-index.json and .plain.html page content from a user-provided domain to build a link graph. This behavior is necessary for the skill's purpose and is performed within the context of the user-scoped audit.
[DATA_EXFILTRATION]: Analysis of the fetch operations and link processing logic shows no patterns of unauthorized data exfiltration or access to sensitive local environment variables, credentials, or system files.
[PROMPT_INJECTION]: The skill includes a dedicated 'External Content Safety' section that explicitly instructs the agent to treat all fetched content as untrusted input. It forbids following redirects to unspecified domains, executing scripts, or interpreting dynamic content from the fetched HTML, which significantly mitigates the risk of indirect prompt injection.

internal-linking