cc-codex-spec-bootstrap

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts located in the .trellis/scripts/ directory to manage task orchestration and context extraction.
  • [COMMAND_EXECUTION]: Orchestrates parallel tasks by executing codex CLI commands that process generated PRDs to fill specification files.
  • [EXTERNAL_DOWNLOADS]: Fetches and installs architectural analysis tools from the npm registry using npm install -g abcoder and npx gitnexus, which are standard package management operations for developer tools.
  • [REMOTE_CODE_EXECUTION]: Configures Model Context Protocol (MCP) servers using npx -y gitnexus mcp to provide the agent with graph-based analysis capabilities from a remote package repository.
  • [PROMPT_INJECTION]: The skill analyzes repository source code (untrusted data) to generate instructions for subsequent tasks. Ingestion points include GitNexus and ABCoder analysis; boundary markers are provided via specific PRD templates and task scoping; capabilities include file writing and agent orchestration; no explicit sanitization of embedded source code instructions is performed.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 10:01 AM