code-intelligence

Warn

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to download and install third-party utilities from external sources including npm (gitnexus) and GitHub (github.com/neur0map/prowl).
  • [REMOTE_CODE_EXECUTION]: MCP server registration via 'npx -y gitnexus@latest mcp' allows the agent to download and execute code from a third-party registry at runtime.
  • [COMMAND_EXECUTION]: Includes multiple shell commands for tool installation and indexing. The 'rename' tool performs direct file system modifications across multiple files.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by indexing codebase content. 1. Ingestion points: Local codebase files processed via tree-sitter (indexing-pipeline.md). 2. Boundary markers: Absent. 3. Capability inventory: 'rename' (file-write), 'cypher' (query-based read), and indexing commands (SKILL.md). 4. Sanitization: Not specified.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 16, 2026, 02:49 PM
Security Audit — agent-trust-hub — code-intelligence