interactive-json-form

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a Bash script template in SKILL.md for deploying generated HTML files to a remote server. This template uses common network tools such as scp and ssh to facilitate file transfer and remote command execution.\n- [PRIVILEGE_ESCALATION]: The provided deployment recipe in SKILL.md utilizes sudo for several operations on the target host, including creating directories within the web server's protected document root, copying files, and adjusting file ownership to match the web server user.\n- [INDIRECT_PROMPT_INJECTION]: The skill contains a vulnerability surface for processing untrusted data.\n
  • Ingestion points: Data enters the skill via the JSON import feature in templates/index.html, which uses a FileReader to ingest external files.\n
  • Boundary markers: The logic does not implement specific boundary markers or instructions to the agent to disregard potential instructions embedded within the ingested JSON content.\n
  • Capability inventory: The skill uses Bash, Write, and Edit tools to generate and manage the interactive content.\n
  • Sanitization: The skill uses JSON.parse() and directly assigns data to element properties. This approach prevents script execution in the browser but does not sanitize the input against malicious instructions meant for the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 02:13 AM