react-bits

Warn

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches a component registry, one-line descriptions, and component source code from https://reactbits.dev via the scripts/rb-add.mjs and scripts/gen-catalog.mjs scripts.\n- [REMOTE_CODE_EXECUTION]: The script scripts/rb-add.mjs downloads code content from an external registry and writes it directly to the local filesystem (defaulting to src/components). This process introduces external, unverifiable code into the local project environment.\n- [COMMAND_EXECUTION]: The script scripts/rb-add.mjs utilizes execFileSync to execute npm install. The specific package names and version ranges are retrieved dynamically from the remote registry at reactbits.dev, allowing an external source to determine which packages are installed and executed on the system.\n- [INDIRECT_PROMPT_INJECTION]: The skill's reliance on external registry data creates a vulnerability surface for indirect injection if the source is compromised.\n
  • Ingestion points: The skill fetches JSON and text data from https://reactbits.dev/r/registry.json and https://reactbits.dev/llms.txt.\n
  • Boundary markers: No specific boundary markers or 'ignore' instructions are applied to the fetched source code or package names.\n
  • Capability inventory: The skill can perform filesystem writes and execute shell commands through the npm CLI.\n
  • Sanitization: The scripts include a basic check to ensure fetched content is not an HTML shell before attempting to parse it as JSON.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 17, 2026, 08:14 AM
Security Audit — agent-trust-hub — react-bits