legal-ai-audit-bundle

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The primary script, scripts/assemble-bundle.mjs, is a zero-dependency Node.js utility that performs file organization, hashing, and manifest generation locally within the agent's environment.
  • [DATA_EXPOSURE]: The skill instructions provide specific guidance on excluding pseudonymization maps (PII) from audit bundles to comply with RODO (GDPR) requirements, promoting secure handling of sensitive legal data.
  • [COMMAND_EXECUTION]: The script uses standard node:fs modules to copy files and create directories based on a provided descriptor. These operations are within the scope of the skill's allowed-tools (Read, Write, Bash).
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data through a JSON descriptor. 1. Ingestion points: descriptor.json read in scripts/assemble-bundle.mjs. 2. Boundary markers: None. 3. Capability inventory: The script performs mkdirSync, copyFileSync, writeFileSync, and readFileSync in scripts/assemble-bundle.mjs. 4. Sanitization: The script uses basename() for destination paths to prevent directory traversal outside the bundle target.
  • [EXTERNAL_DOWNLOADS]: The skill references established patterns from Microsoft and ICME for audit logs as roadmap inspiration but does not download or execute remote code at runtime.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 10:25 AM