kamis-food-price
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The Python script
scripts/run_kamis.pyaccesses a local configuration file at~/.config/k-skill/secrets.envto retrieve theKSKILL_KAMIS_API_KEY. This key is then transmitted via network requests to the vendor proxy (k-skill-proxy.nomadamas.org) or the official upstream API (www.kamis.or.kr). - [EXTERNAL_DOWNLOADS]: The skill instructions in
SKILL.mddirect the user to runnpx -y @nomadamas/k-skill@0, which downloads and executes code from the npm registry. - [COMMAND_EXECUTION]: The skill utilizes shell commands (
npx,curl, andpython) to fetch instructions, manage dependencies, and execute the price lookup logic. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external APIs, which serves as a potential vector for indirect prompt injection if the API source were compromised.
- Ingestion points: Data is ingested from
k-skill-proxy.nomadamas.organdwww.kamis.or.krwithinscripts/run_kamis.py. - Boundary markers: No explicit delimiters or instructions to ignore embedded content are used when processing the API payload.
- Capability inventory: The skill possesses network access capabilities and the ability to read local configuration files.
- Sanitization: API responses are printed as raw text or JSON without specific sanitization or filtering to prevent the agent from interpreting embedded instructions.
Audit Metadata