s2b-notice-search
Warn
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to execute
npx -y @nomadamas/k-skill@0to fetch instructions, list files, and perform updates. This creates a dependency on an external NPM package from a non-trusted source. - [REMOTE_CODE_EXECUTION]: The use of
npxto run remote packages (@nomadamas/k-skill) constitutes remote code execution, as the package content is fetched from the NPM registry and executed locally at runtime. - [COMMAND_EXECUTION]: The skill provides specific bash commands for the agent to execute as part of its setup and operation, including package management and instruction retrieval.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to parse HTML content from the S2B procurement portal (
www.s2b.kr). This represents a vulnerability surface where malicious content on the target site could potentially influence the agent's behavior during the parsing process. - Ingestion points: HTML results and search forms from
https://www.s2b.kr/S2BNCustomer/tcmo001.do. - Boundary markers: None identified in the instruction files to separate untrusted web content from agent instructions.
- Capability inventory: Browser automation (Aside Browser, CDP), form submission, and execution of logic from the
s2b-notice-searchpackage. - Sanitization: No explicit sanitization or validation of the parsed HTML content is mentioned before it is processed by the agent.
Audit Metadata