lov-better-project-structure

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill's workflow includes running actual build and related tests (Workflow step 5 in SKILL.md). This involves executing scripts defined within the target project's environment, such as build commands or test suites.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external project data which could theoretically contain malicious instructions.
  • Ingestion points: The skill reads project manifest files, entry points, build configurations, and Git status from the target project (SKILL.md).
  • Boundary markers: The execution boundary in SKILL.md explicitly separates analysis and preview (read-only) from modification and submission tasks to reduce accidental execution.
  • Capability inventory: The skill can modify files, update imports/references, and execute shell commands for builds and tests (SKILL.md).
  • Sanitization: The scripts/profile_store.py script includes a validation check (target_parts) that prevents the persistence of keys containing sensitive substrings such as 'password', 'token', or 'api_key'.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 09:45 AM
Security Audit — agent-trust-hub — lov-better-project-structure