court-payment-order-assistant

Warn

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill employs a concealment pattern by directing the agent to fetch its "primary instructions" via a CLI tool (npx -y @nomadamas/k-skill@0 instruct) instead of relying on the provided instruction.md. This dynamic loading prevents static analysis of the agent's actual operational rules, as the CLI output can change independently of the skill repository.
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill requires the execution of npx -y @nomadamas/k-skill@0, which downloads and executes code from the npm registry. While the package belongs to the skill's author, the use of npx for runtime instruction delivery is a high-risk pattern.
  • [DYNAMIC_EXECUTION]: The instruction.md file contains code snippets that utilize require("court-payment-order-assistant"). This indicates a dependency on external, dynamically loaded modules for core functionality, such as generating legal drafts and handling browser interactions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 20, 2026, 07:36 AM
Security Audit — agent-trust-hub — court-payment-order-assistant