DannyMac180/sol-advisor

Codex-native architect orchestration with Luna and Terra implementation lanes and mandatory fresh Sol review.

What it solves

Sol Advisor is designed to streamline software delivery by automating the planning, implementation, and verification of features. It removes the need for the user to manually manage different AI agents or "lanes" by allowing a primary agent (Sol / High) to own the entire lifecycle of a task from goal setting to final acceptance.

How it works

The system uses a "capability-routed" workflow within the Codex CLI or ChatGPT desktop app. A primary agent, Sol / High, evaluates the risk and complexity of a task to select a delivery route:

  • Solo: The primary agent handles everything (planning, implementation, testing, and review).
  • Delegate: The work is delegated to a specialized agent (Luna / Max or Terra / High) for execution, while the primary agent verifies the result.
  • Audit: The primary agent implements the work, and a separate, fresh Sol / High agent performs an independent review.
  • Full: A high-risk exception where one agent implements and another reviews, with root verification.

Who it’s for

Developers using the Codex CLI or ChatGPT desktop app who want an agentic workflow for software engineering tasks where the agent manages its own routing and quality assurance.

Highlights

  • Risk-Gated Routing: Automatically selects the delivery mode based on risk rationale before calling task tools.
  • Automated Verification: The primary agent maintains architecture and decomposition while ensuring all auxiliary work is verified.
  • Fail-Closed Installation: Includes a companion installer that verifies role files to ensure they are not modified or unsafe.
  • Orchestration-First: Uses a specific orchestration prompt to trigger the capability-routed delivery process.

Related

  • Project
    viettran-edgeAI/codex_workflow

    A token-efficient agent orchestration workflow for Codex that uses a tiered routing system and specialized sub-agents to manage complex software projects.

  • Project
    phodal/routa

    A workspace-first multi-agent coordination platform for software delivery that uses a Kanban-based routing system to manage goals, tasks, and evidence.

  • Project
    michaelshimeles/skills

    A collection of specialized agent skills for Claude Code that automate complex workflows like visual testing, architectural enforcement, and PR refinement.

  • Project
    Cjbuilds/Codex-Orchestration

    Codex Orchestration is a Codex plug‑in that lets you attach other LLMs (Claude Fable 5, Opus 5, OpenRouter models, etc.) to a task and assign them specific roles—Planner, Advisor, Designer, Executor. Codex stays the top‑level orchestrator, iterating plans, reviewing them, optionally designing UX artefacts, and finally executing code. The plug‑in is installed via the Codex marketplace, configured with skill prompts inside a Codex chat, and supports role‑specific effort levels. It aims to improve plan quality, enable parallel work, and reduce premium‑model usage. MIT‑licensed.

  • Project
    umputun/ralphex

    A standalone CLI tool that orchestrates Claude Code or codex to autonomously execute implementation plans, featuring a multi-phase AI code review pipeline.