modernize-ecmascript

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches an index of ECMAScript proposals and technical metadata from https://bosens-china.github.io/tc39-atlas/llms.txt. This information is used to guide the agent in determining which modern language features are compatible with the current project's build chain.- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external sources (such as TC39 proposal descriptions and project-specific configuration files) which could influence the agent's code modification and command execution behavior.
  • Ingestion points: Fetches data from external URLs (TC39 Atlas) and reads local project files (package.json, build configs).
  • Boundary markers: The skill lacks explicit delimiters for external content but instructs the agent to verify implementation status against official repositories.
  • Capability inventory: The agent is authorized to modify source code and execute project commands (e.g., build, test, lint) based on the analysis of these external inputs.
  • Sanitization: The instructions mandate user confirmation for significant changes like adding new dependencies, polyfills, or changing build targets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 04:43 AM
Security Audit — agent-trust-hub — modernize-ecmascript