lh-notice-search
Fail
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill instructs the agent to run
npx -y @nomadamas/k-skill@0 instruct lh-notice-searchas a required first step. This pattern fetches and executes a remote JavaScript package from a public registry at runtime, which is a high-risk pattern for arbitrary code execution. - [EXTERNAL_DOWNLOADS]: Core instructions and helper files are downloaded from a remote repository (github.com/NomaDamas/k-skill) and an external npm package, bypassing local file review and static analysis.
- [COMMAND_EXECUTION]: The skill uses shell commands like
npxandcurlto interact with external services (k-skill-proxy.nomadamas.org), which are managed by a third party and not on the established list of trusted or well-known services. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external API responses via a proxy. This creates a surface for indirect prompt injection if the upstream source or proxy is compromised. Evidence: Ingestion point at /v1/lh-notice/search; no explicit boundary markers or sanitization are defined for the received payloads.
Recommendations
- AI detected serious security threats
Audit Metadata