incu-way-init

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes repository files like package.json and source code to automate setup, which exposes it to malicious instructions embedded in those files.
  • Ingestion points: Survey of repository root (Phase 0), extraction of commands from manifests (Phase 2.1), and architectural review (Phase 3).
  • Boundary markers: Absent; the skill does not use delimiters to wrap or isolate the data read from the repository.
  • Capability inventory: Writing configuration files, managing git branches/worktrees, and triggering other project-specific skills.
  • Sanitization: Absent; data read from the repository is directly interpolated into project templates without validation.
  • [COMMAND_EXECUTION]: The skill utilizes shell commands (git, ls, mkdir) to inspect the repository state and initialize the working environment. All high-impact operations are protected by mandatory user confirmation gates.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 01:49 PM
Security Audit — agent-trust-hub — incu-way-init