using-git-worktrees

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a framework for processing external web content and historical integrations while possessing tools for terminal execution and filesystem persistence, creating a vulnerability surface.
  • Ingestion points: The skill ingests untrusted data from web content, external sources via BrowserOS, and historical prompt integrations as noted in the compatibility section.
  • Boundary markers: There are no explicit technical delimiters or specific prompt boundary markers defined in the instructions, though the agent is directed to treat external data as hostile.
  • Capability inventory: The skill utilizes the ExecutionProxy Interface (e.g., rtk commands) for shell execution, the MemoryStore Interface for writing markdown artifacts to the filesystem, and various MCP tools for context retrieval and reasoning.
  • Sanitization: The instructions mandate redacting secrets and PII before sharing context with subagents, but lack technical sanitization or escaping mechanisms for data interpolated into shell commands or prompt context.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 07:35 AM
Security Audit — agent-trust-hub — using-git-worktrees