skills/realwigu/claude-skills/init/Gen Agent Trust Hub

init

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies the project framework and language by reading local files, creating a vulnerability surface where a malicious project configuration could influence the agent's actions.
  • Ingestion points: Reads package.json and tsconfig.json in the project root to determine setup parameters.
  • Boundary markers: Does not implement specific delimiters or instructions to ignore potential commands within the ingested project files.
  • Capability inventory: The skill can perform file system writes, install packages via npm, and execute shell commands (npx playwright test).
  • Sanitization: No validation or sanitization of the content extracted from project files is performed before it influences the setup logic.
  • [COMMAND_EXECUTION]: The skill executes shell commands to provision the environment and run tests.
  • Evidence: Uses npm init playwright@latest for initialization and npx playwright test for verification in Steps 2 and 9.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 12:32 PM
Security Audit — agent-trust-hub — init