ln-730-devops-setup

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches supporting reference files from the author's official GitHub repository (levnikolaevich/claude-code-skills) when they are missing from the local environment.
  • [COMMAND_EXECUTION]: Executes system commands including docker --version, node -v, and docker-compose config to perform environment validation and syntax checks on generated files.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes untrusted data from project files to automatically detect the technology stack.
  • Ingestion points: Reads metadata and configuration strings from package.json, requirements.txt, and .env.example (SKILL.md).
  • Boundary markers: No explicit delimiters or instructions to ignore embedded content are used when interpolating detected data into worker instructions.
  • Capability inventory: The skill executes shell commands (docker-compose config) and coordinates file-writing tasks across multiple generator workers.
  • Sanitization: No explicit sanitization, validation, or escaping of the content extracted from project files is documented.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 05:00 PM