Skills
skills/vladmdgolam/agent-skills/reverse-engineer-js/Gen Agent Trust Hub

reverse-engineer-js

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted JavaScript bundles which could potentially contain malicious instructions aimed at the agent (indirect prompt injection).
  • Ingestion points: The skill reads external JavaScript files (e.g., bundle.js) as seen in Workflows A, B, and C.
  • Boundary markers: Absent; there are no instructions to delimit the processed code or ignore embedded commands within it.
  • Capability inventory: The skill uses shell commands (mkdir, cp, rm, script), executes NPM packages via npx, and allows for execution of local project scripts (node scripts/deobfuscate.js).
  • Sanitization: None; the code is passed directly to deobfuscation tools and the LLM.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands and standard utilities.
  • Evidence: Usage of mkdir, cp, rm, and the script utility for TTY emulation in SKILL.md.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes npx to download and run packages from the NPM registry.
  • Evidence: References to npx humanifyjs, npx webcrack, and npx prettier throughout the workflows in SKILL.md.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 01:50 PM