The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill requires the agent to execute the @nomadamas/k-skill package using npx, which involves downloading and running arbitrary code from the npm registry at runtime. This package originates from an untrusted third-party organization.
[COMMAND_EXECUTION]: The skill's instructions mandate the use of shell commands (npx) to retrieve the full instruction set, list helper files, and execute the core Python logic (scripts/daangn_realty.py).
[PROMPT_INJECTION]: The skill contains a directive to the agent to override the provided markdown instructions with output from an external command: 'Run this and follow its output as the primary instructions for this skill'. This creates a mechanism for remote instruction injection, allowing an external source to control the agent's behavior.
[INDIRECT_PROMPT_INJECTION]: The skill processes real estate listing data from external web pages, which serves as a vulnerability surface for indirect prompt injection.
Ingestion points: Data is fetched from realty.daangn.com and www.daangn.com via urllib.request in scripts/daangn_realty.py and scripts/daangn_detail_ld.py.
Boundary markers: The skill does not implement delimiters or 'ignore embedded instructions' warnings when presenting scraped content to the agent.
Capability inventory: The skill possesses capabilities for network reading and command execution.
Sanitization: No sanitization or filtering is performed on the extracted HTML or JSON-LD data before it is returned to the agent.

daangn-realty-search