deveco-requirement-development

Warn

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructs users to configure an MCP server using npx -y @deveco-codegenie/mcp@beta in the common configuration rules. This command automatically downloads and executes code from the npm registry without manual verification of the package content.
  • [COMMAND_EXECUTION]: The development workflow relies on several CLI tools such as check_ets_files, build_project, and start_app to perform static analysis, compilation, and application deployment on the user's local system.
  • [EXTERNAL_DOWNLOADS]: The skill references external design platforms (Figma, MasterGo, Sketch) and a third-party GitHub repository (https://github.com/open-deveco/deveco-toolbox) for project tooling and requirement extraction.
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface:
  • Ingestion points: The skill ingests DSL data from external design tool URLs and reads existing project files (e.g., package.json, source code) during the requirement development phase.
  • Boundary markers: No specific delimiters or instructions are provided to separate potentially untrusted design data from the agent's core instructions.
  • Capability inventory: The agent has the ability to write multiple files (PRDs, investigations, code), perform codebase searches, and execute build/deploy commands.
  • Sanitization: The instructions do not define any validation or sanitization steps for the data retrieved from external design MCPs before using it for PRD or code generation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 22, 2026, 10:16 AM
Security Audit — agent-trust-hub — deveco-requirement-development