kiwi-rs-assistant
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill describes a library bootstrapping process using
curlandtar(orpowershellon Windows) to manage binary runtime assets as mentioned inSKILL.mdandreferences/troubleshooting.md. - [COMMAND_EXECUTION]: Instructions guide the user to execute
cargo runcommands to validate and test the generated Rust code snippets. - [PROMPT_INJECTION]: An indirect prompt injection surface is present in the data processing flow.
- Ingestion points: Korean text input provided by the user for tokenization and analysis tasks in
SKILL.mdandreferences/task-to-api.md. - Boundary markers: None explicitly mentioned in instructions; the generated code handles input as standard Rust string variables.
- Capability inventory: Capabilities are limited to NLP processing and terminal output; no network exfiltration or file-system writing based on user input is generated.
- Sanitization: No sanitization or filtering of input text is performed before it is passed to the NLP library.
Audit Metadata