rails-project-discovery

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads and processes local project files (Gemfile.lock, config/application.rb, config/database.yml, config/routes.rb) and user-provided project details to route to other skills. Maliciously crafted project files could potentially influence agent behavior.
  • Ingestion points: Project configuration files and developer interview responses (SKILL.md, "I'm joining an existing Rails project" section).
  • Boundary markers: No specific delimiters or safety warnings are instructed when reading external file content.
  • Capability inventory: The skill suggests executing local shell commands (rails new, bundle add, rails g) and determines which subsequent skills are loaded into the agent's context.
  • Sanitization: No explicit sanitization or validation of the ingested file content is performed before logic branching.
  • [COMMAND_EXECUTION]: The skill suggests executing shell commands such as rails new, bundle add, and bin/rails generate authentication. These are standard Rails development commands and are intended to be executed by the user/agent in the project directory.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 03:48 AM
Security Audit — agent-trust-hub — rails-project-discovery