Skills
Navigate
skills/growthx-club/build-sprint-skills/agentation/Gen Agent Trust Hub

agentation

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted browser-based annotations into the agent context, creating a potential injection surface.
  • Ingestion points: The agentation_get_all_pending tool reads annotations from the browser toolbar (SKILL.md).
  • Boundary markers: No delimiters or "ignore instructions" warnings are specified for the ingested annotation data.
  • Capability inventory: The skill can install packages, execute shell commands, and modify application layout files.
  • Sanitization: No sanitization or validation of the retrieved browser feedback is performed.
  • [REMOTE_CODE_EXECUTION]: The skill downloads and executes external code from the npm registry during setup.
  • Evidence: Commands such as npm install agentation, npx -y agentation-mcp server, and npx agentation-mcp doctor download and run code at runtime.
  • [COMMAND_EXECUTION]: The skill executes various shell commands to configure agent capabilities and diagnose services.
  • Evidence: The use of claude mcp add and npx add-mcp for agent configuration, and npx agentation-mcp doctor for service health checks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 09:59 AM