research-lookup
Fail
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The SKILL.md file includes instructions for the user to download and execute a remote script via 'curl -fsSL https://parallel.ai/install.sh | bash'. This pattern allows for the execution of arbitrary, unverified code from a source outside the trusted list.
- [EXTERNAL_DOWNLOADS]: The skill requires downloading external command-line tools and scripts from parallel.ai.
- [PROMPT_INJECTION]: The research lookup logic in research_lookup.py and scripts/research_lookup.py fetches results from external search APIs (Parallel and Perplexity) and integrates them into the agent's context. The lack of sanitization or explicit boundary markers for this untrusted external data makes the skill vulnerable to indirect prompt injection.
- [COMMAND_EXECUTION]: The script scripts/generate_schematic.py uses subprocess.run() to execute the internal script scripts/generate_schematic_ai.py during the schematic generation process.
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