ollama-setup

Fail

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches installation scripts and binaries from the official Ollama domain.
  • Evidence: SKILL.md and references/skill-workflow.md reference https://ollama.com/install.sh.
  • Evidence: references/examples.md references binary downloads from https://ollama.com/download/ollama-linux-amd64.
  • [REMOTE_CODE_EXECUTION]: The skill executes a remote installation script via a shell pipe, which is the official deployment method for Linux.
  • Evidence: SKILL.md contains the command 'curl -fsSL https://ollama.com/install.sh | sh'.
  • [COMMAND_EXECUTION]: The skill utilizes elevated privileges to install binaries and manage system services.
  • Evidence: SKILL.md and references/examples.md contain 'sudo systemctl' and 'sudo cp' commands for service configuration.
  • [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface through its automated code review example.
  • Ingestion points: references/examples.md contains a Node.js CI script that processes arbitrary git diff strings.
  • Boundary markers: The implementation lacks explicit delimiters or instructions to ignore commands embedded in the processed data, relying only on a JSON format request.
  • Capability inventory: The skill possesses broad Bash, Write, and Read capabilities, which could be abused if an injection succeeds.
  • Sanitization: There is no evidence of sanitization or escaping of the input diff content before it is interpolated into the prompt.
Recommendations
  • HIGH: Downloads and executes remote code from: https://ollama.com/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
May 9, 2026, 01:10 AM