commercetools-spec-driven-development

Pass

Audited by Gen Agent Trust Hub on Sep 12, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The scripts/docs-search.mjs script performs GET requests to https://docs.commercetools.com. This network operation targets an official vendor domain to provide relevant documentation as grounding context for the agent and does not involve downloading or executing untrusted code.
  • [COMMAND_EXECUTION]: The scripts/setup.mjs utility performs local file system operations to modify project configuration files (e.g., in .specify/ or openspec/). The script is restricted to specific hardcoded targets and uses marker-delimited text substitution to ensure that changes are idempotent and reversible.
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes an indirect injection surface by modifying development templates (such as plan-template.md) that are subsequently read by agents. These instructions guide the agent to load specific commercetools-* skills and verify actions using the commercetools-knowledge MCP. The process is managed using clear boundary markers and lacks unsafe interpolation of untrusted data, maintaining a safe environment for agent operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 12, 2026, 10:02 PM
Security Audit — agent-trust-hub — commercetools-spec-driven-development