sharenuri-facility-search
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill's primary documentation (SKILL.md and instruction.md) directs the agent and user to execute commands using
npx -y @nomadamas/k-skill@0. This results in the automatic download and execution of code from an external NPM package registry. - [COMMAND_EXECUTION]: The skill instructions depend on the execution of shell commands through
npxto perform core tasks such as displaying instructions, listing bundled files, and updating the skill components. - [INDIRECT_PROMPT_INJECTION]: The
scripts/sharenuri_search.pyscript fetches and parses HTML content from the external government portaleshare.go.kr. While the script performs basic sanitization by stripping HTML tags, the extracted text (such as facility names and descriptions) is directly presented to the agent. This creates a surface where malicious instructions embedded in the source data could potentially influence the agent's behavior. - Ingestion points: Search results and detail pages fetched from
https://www.eshare.go.krinscripts/sharenuri_search.py. - Boundary markers: None identified between external data and agent instructions.
- Capability inventory: Network access via
urllib.requestto communicate with the external portal. - Sanitization: Employs
_strip_htmlto remove tags and normalize whitespace before returning data to the agent.
Audit Metadata