powerpoint-automation

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Several Python and Node.js scripts utilize subprocess modules (e.g., subprocess.run and execSync) to perform functional tasks. These include retrieving repository metadata via Git commands and opening generated PowerPoint files for user review. The commands are generally static or utilize sanitized input.
  • [EXTERNAL_DOWNLOADS]: The skill implements logic to download images from user-provided URLs to embed them into slides. This is handled through standard libraries like urllib.request and Node's https module. While fetching external content is inherently risky, the implementation is scoped to the skill's purpose of presentation creation.
  • [INDIRECT_PROMPT_INJECTION]: The skill's architecture involves ingesting untrusted data from web articles and existing PowerPoint files to generate new content. There is a potential surface for indirect prompt injection as external text is parsed and re-processed by the agent without explicit boundary markers or isolation instructions in the intermediate JSON representation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 03:31 AM
Security Audit — agent-trust-hub — powerpoint-automation