ucoz-provisioning-skill

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a browser automation runner located in the browser-runner/ directory. The main script ucoz-provision.cjs is executed via Node.js to interact with the uCoz control panel.
  • [DYNAMIC_EXECUTION]: The compatibility shim provision-new-site.mjs uses node:child_process.spawnSync to execute the main provisioning runner. It includes a safety mechanism that filters environment variables, ensuring only necessary parameters (like UCOZ_* prefixes) are passed to the child process, which prevents the leakage of unrelated system secrets.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the Playwright automation framework and its browser binaries. It also recommends running the official ucoz-mcp server using npx -y ucoz-mcp@latest. These are standard requirements for the skill's stated purpose of site provisioning and MCP integration.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it reads domain zones, theme configuration cards, and module lists directly from the uCoz web interface.
  • Ingestion points: Browser DOM extraction in ucoz-provision.cjs (functions getDomainZones, getPresetCards, listPanelModules).
  • Boundary markers: The skill instructions in SKILL.md mandate that the agent must clarify and confirm all extracted parameters with the user in chat before proceeding with execution.
  • Capability inventory: The skill has capabilities for command execution via spawnSync, file writing for results, and authenticated network requests via automated browser sessions.
  • Sanitization: The runner uses specific CSS selectors and string normalization to extract data, reducing the risk of processing unintended content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 02:45 PM