ln-781-build-verifier

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches missing reference files from the author's GitHub repository at https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path}. This is a vendor-owned resource but involves fetching configuration from a remote source if local files are missing.\n- [COMMAND_EXECUTION]: Executes a variety of build and dependency restoration tools including npm, yarn, pnpm, dotnet, pip, poetry, go, cargo, mvn, and gradle based on detected project types in the local environment.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by processing external data and using it to drive execution.\n
  • Ingestion points: Reads project-specific files such as package.json, pyproject.toml, Cargo.toml, pom.xml, and build.gradle (SKILL.md) to determine project types and build steps.\n
  • Boundary markers: No specific delimiters or instructions to ignore embedded malicious content (like instructions within package.json scripts) are present in the skill's workflow.\n
  • Capability inventory: Executes shell commands and build processes through the Monitor or Bash tools, providing a path for execution of malicious scripts defined in ingested files.\n
  • Sanitization: Lacks explicit validation or sanitization for the content of the ingested configuration files before executing the related build and restoration commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 08:42 AM