parallel-web-search

Fail

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill contains a setup instruction to download and execute a script from https://parallel.ai/install.sh using curl -fsSL ... | bash. This allows arbitrary remote code execution from an unverified source.
  • [COMMAND_EXECUTION]: The skill executes shell commands using $ARGUMENTS in parallel-cli search "$ARGUMENTS". This presents a risk of command injection if the input contains shell metacharacters that escape the double-quote encapsulation.
  • [EXTERNAL_DOWNLOADS]: The skill attempts to install the parallel-web-tools[cli] package from an external repository using pipx.
  • [CREDENTIALS_UNSAFE]: The instructions guide the user to set sensitive API keys via environment variables (export PARALLEL_API_KEY="your-key"), which can lead to credential exposure in shell history or environment logging.
  • [PROMPT_INJECTION]: The skill processes untrusted content from the web and synthesizes it into a response without safety delimiters.
  • Ingestion points: Search results retrieved from the web via parallel-cli (SKILL.md).
  • Boundary markers: No delimiters or "ignore instructions" warnings are defined for processing search results (SKILL.md).
  • Capability inventory: The skill has access to the Bash tool and file system via /tmp (SKILL.md).
  • Sanitization: No sanitization or validation of the retrieved web content is performed before interpolation into the final response (SKILL.md).
Recommendations
  • HIGH: Downloads and executes remote code from: https://parallel.ai/install.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 27, 2026, 06:42 PM