plannotator-visual-explainer

Warn

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires installing a third-party extension nicobailon/visual-explainer via npx skills add. This source is not from a trusted organization.
  • [COMMAND_EXECUTION]: Instructs the use of npx skills add -g --yes, which installs extensions globally and bypasses manual confirmation prompts.
  • [REMOTE_CODE_EXECUTION]: Generated visualizations embed a script that executes code from a remote CDN (cdn.jsdelivr.net) for the @pierre/diffs package.
  • [DYNAMIC_EXECUTION]: The test suite SKILL.test.ts uses new Function() and Bun.spawn with the -e flag to run dynamically constructed scripts.
  • [INDIRECT_PROMPT_INJECTION]: The skill constructs HTML reports by interpolating untrusted inputs without sanitization.
  • Ingestion points: Processes data from PR metadata, code diffs, and user briefs.
  • Boundary markers: No explicit delimiters are used to protect against malicious data patterns.
  • Capability inventory: Uses plannotator annotate and file system metadata access.
  • Sanitization: No evidence of input escaping is present in the HTML component definitions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 17, 2026, 03:51 PM
Security Audit — agent-trust-hub — plannotator-visual-explainer