nextreme-flowcharts

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the agent to execute local scripts (scripts/generate_pdf.py and scripts/verify_pdf.py) to transform generated HTML into PDF format and to create visual verification images. These operations involve running subprocesses on the host environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts natural language descriptions of logic, hierarchies, and learning paths from users and interpolates them into HTML/SVG templates, creating a surface for injection attacks.
  • Ingestion points: User-provided descriptions of flowcharts, roadmaps, and diagrams described in the skill's trigger conditions.
  • Boundary markers: Absent. There are no instructions or markers to distinguish user data from the HTML/SVG shell structure or to warn the agent about ignoring potential instructions within user input.
  • Capability inventory: The skill uses scripts/generate_pdf.py (which invokes rendering backends like Playwright or WeasyPrint) and scripts/verify_pdf.py to write files to the disk and potentially execute further subprocesses.
  • Sanitization: Absent. The skill provides no guidelines or requirements for sanitizing or escaping user-supplied text before it is embedded into the HTML source.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 12:59 PM
Security Audit — agent-trust-hub — nextreme-flowcharts