g2b-sanctioned-supplier
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions in
SKILL.mdsuggest usingnpx -y @nomadamas/k-skill@0to fetch instructions and updates. This executes a package within the author's (@nomadamas) NPM namespace, which is consistent with the vendor's established deployment pattern. - [COMMAND_EXECUTION]: The skill utilizes a Python script
scripts/g2b_sanctioned_supplier.pythat makes network requests using the standardurlliblibrary to a vendor-controlled domain (k-skill-proxy.nomadamas.org). The script handles business registration numbers to lookup public information. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data retrieved from a procurement database, creating a potential surface for indirect injection.
- Ingestion points: Untrusted data enters the agent context through the JSON output of
scripts/g2b_sanctioned_supplier.pyvia an HTTP GET request to the proxy. - Boundary markers: There are no explicit boundary markers or safety delimiters surrounding the interpolated JSON output in the instructions.
- Capability inventory: The skill possesses capabilities for network operations (querying the proxy) and local command execution (running the helper script).
- Sanitization: The Python helper performs basic JSON validation and UTF-8 decoding but does not sanitize the content of the fields returned by the upstream API.
Audit Metadata