graphify-codebase-architect

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install an external dependency 'graphifyy[mcp]' using the 'uv' tool, which involves downloading code from a public package registry.\n- [COMMAND_EXECUTION]: The setup and workflow require executing shell commands such as 'graphify .' and 'python -m graphify.serve' to parse the codebase and host the MCP server.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes entire codebases to extract AST structural data, creating a surface for indirect prompt injection from untrusted source code.\n
  • Ingestion points: Codebase files parsed by the 'graphify' tool as specified in SKILL.md.\n
  • Boundary markers: The skill contains explicit instructions to maintain zero exposed secrets and ensures visual outputs are local-first.\n
  • Capability inventory: The agent can execute shell commands, create files (graph.json, graph.html), and provide query tools via MCP.\n
  • Sanitization: The instructions include a quality checklist to verify that all credentials have been scrubbed from generated reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 04:56 AM
Security Audit — agent-trust-hub — graphify-codebase-architect