The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill connects to https://jobdanmark.dk to fetch job data, categories, and detailed job descriptions. These network operations are necessary for the skill's intended purpose.
[PROMPT_INJECTION]: The detail command retrieves and processes job descriptions from external job postings on jobdanmark.dk. This data is untrusted and represents a surface for Indirect Prompt Injection, where a malicious listing could attempt to influence the agent's behavior.
Ingestion points: cli/src/commands/detail.ts fetches HTML content from the job portal.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used when presenting the external description to the agent.
Capability inventory: The skill allows network requests and CLI execution, but it does not execute or evaluate the ingested text content.
Sanitization: The skill uses node-html-parser to extract plain text from HTML elements, which provides basic sanitization against script tags but does not prevent natural language instructions.

jobdanmark-search