pp-dreaming

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the dreaming-pp-cli tool via the NPM registry using the command: npx -y @mvanhorn/printing-press-library install dreaming --cli-only.
  • [REMOTE_CODE_EXECUTION]: Using npx to install the library involves the automated download and execution of code from the remote NPM registry.
  • [DATA_EXFILTRATION]: The skill exposes an output delivery mechanism via the --deliver webhook: flag, which enables the routing of command results (potentially containing user account details or private transcripts) to arbitrary external endpoints.
  • [PROMPT_INJECTION]: The skill processes untrusted external data, which creates a surface for indirect prompt injection attacks.
  • Ingestion points: The skill ingests data from external CSV files (external import), YouTube video metadata (inspect-external-video), and video transcripts (transcript and concordance commands).
  • Boundary markers: No specific boundary markers or instructions to ignore embedded directions within the ingested content are defined in the instructions.
  • Capability inventory: The skill possesses the ability to read local files, execute shell commands via the CLI binary, and perform network POST requests through its webhook delivery feature.
  • Sanitization: There is no evidence of content sanitization or validation for the transcripts or imported CSV data before it is presented to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 08:24 PM
Security Audit — agent-trust-hub — pp-dreaming