readme

Warn

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill instructs the agent to access and explore highly sensitive file paths to understand the project configuration. These include config/master.key, config/credentials.yml.enc, and environment variable files like .env.example or .env. Reading the Rails master key or active environment secrets brings the application's full production credentials into the agent's context, creating a significant data exposure risk.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it performs 'Deep Codebase Exploration' by reading arbitrary files from the project root. A malicious file within the codebase (e.g., a README, schema file, or code comment) could contain hidden instructions that manipulate the agent's behavior during the generation process.
  • Ingestion points: Root directory structure, configuration files (config/*.rb, .env), database schemas (db/schema.rb), and dependency manifests (Gemfile, package.json).
  • Boundary markers: Absent; the skill does not use delimiters or instructions to ignore embedded commands in the files it reads.
  • Capability inventory: File-write access to create README.md in the project root.
  • Sanitization: Absent; codebase content is processed directly to generate documentation templates.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 14, 2026, 02:09 PM