10x-cli-setup

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches a README file from the vendor's official GitHub repository (https://raw.githubusercontent.com/przeprogramowani/10x-cli/refs/heads/master/README.md). This is a standard practice for the author to ensure users receive up-to-date installation and configuration instructions.
  • [COMMAND_EXECUTION]: The skill uses shell commands to check the current environment (10x --version) and executes setup, authentication, and diagnostic commands based on the content of the retrieved README. All described behaviors align with the primary purpose of setting up a development tool.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: While the skill handles authentication via a magic-link email, it specifies that this process is interactive and should be handled by the user, avoiding the exposure of credentials to the agent's logs or external endpoints.
  • [INDIRECT_PROMPT_INJECTION]: The skill relies on external content (the README) to drive its logic, which presents an attack surface for indirect prompt injection.
  • Ingestion points: README content fetched via WebFetch or curl from GitHub (SKILL.md).
  • Boundary markers: None explicitly defined between the instructions and fetched data.
  • Capability inventory: Subprocess execution for version checks, installations, and tool configuration.
  • Sanitization: None specified for the README content.
  • However, because the content source is the official repository belonging to the skill's author ('przeprogramowani'), the risk is considered within the trust scope of the vendor.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 02:40 PM