skills/zuplo/tools/zuplo-cli/Gen Agent Trust Hub

zuplo-cli

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents the execution of various CLI commands via npx zuplo to perform project operations, such as starting development servers, managing tunnels, and deploying code to the Zuplo platform.
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies surfaces where external data is ingested into the agent's workflow, creating an indirect prompt injection risk.
  • Ingestion points: The agent is instructed to merge OpenAPI specifications from user-provided URLs (zuplo openapi merge -s <URL>) and analyze test results from remote endpoints (zuplo test --endpoint <URL>).
  • Boundary markers: No explicit delimiters or instructions to ignore embedded prompts within external files are defined in the instructions.
  • Capability inventory: The agent has the capability to write local configuration files, perform network requests, and execute shell commands using the Zuplo CLI.
  • Sanitization: No specific validation or sanitization routines for the external OpenAPI data or test responses are mentioned.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch OpenAPI specifications from remote URLs and utilizes the zuplo and create-zuplo-api packages from the NPM registry. These resources originate from the vendor's official distribution channels.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 07:22 PM