product-design-research

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a significant surface for indirect prompt injection because it is designed to ingest and process data from a wide range of external sources such as Reddit, X (Twitter), and GitHub, as well as internal platforms like Slack and Jira. An attacker could place malicious instructions in these sources to influence the agent's behavior.
  • Ingestion points: Workflow steps in SKILL.md specify searching public and internal sources.
  • Boundary markers: No specific boundary markers or instructions to ignore embedded commands are present for the research data.
  • Capability inventory: The skill has the ability to execute shell commands (node, npm install) and interact with a browser, which could be exploited if the agent follows instructions hidden in the researched data.
  • Sanitization: No specific sanitization or filtering logic is described for the retrieved content.
  • [COMMAND_EXECUTION]: The prototyping workflow involves executing shell commands to set up the environment.
  • Evidence: references/local-prototype-preflight.md includes instructions to run a node-based bootstrap script and npm install.
  • [EXTERNAL_DOWNLOADS]: The use of npm install indicates the skill downloads external dependencies from the npm registry during the local prototyping phase.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 06:05 PM
Security Audit — agent-trust-hub — product-design-research