improve-threejs

Warn

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructs the agent to run npx react-doctor@latest, which downloads and executes code from the NPM registry at runtime. This introduces a supply chain risk as the tool is not from a trusted vendor list.\n- [EXTERNAL_DOWNLOADS]: The skill fetches markdown-based fixing rules from an external URL (https://www.react.doctor/prompts/rules/...). This creates an external dependency on content that could be changed or compromised.\n- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by ingesting and acting upon instructions from external markdown files.\n
  • Ingestion points: Remote rule files fetched from the react.doctor domain (SKILL.md Step 5).\n
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to treat this external content as untrusted data.\n
  • Capability inventory: The skill has permissions to modify the local codebase and execute shell commands via npx.\n
  • Sanitization: The skill lacks any verification or sanitization steps for the content fetched from the remote server before it is used to guide code modifications.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 17, 2026, 05:18 PM
Security Audit — agent-trust-hub — improve-threejs