maxritter/pilot-shell

Professional context and harness engineering for Claude Code and OpenAI Codex. Build production-grade software with spec-driven development, TDD, persistent memory, quality gates, code intelligence, human oversight, and end-to-end verification.

What it solves

Coding agents like Claude Code and Codex write code quickly, but they often lack the durable context, quality control, and verification needed for production-grade software. Pilot Shell provides a professional engineering harness that wraps around these agents to ensure implementation is disciplined, verified, and grounded in persistent project knowledge.

How it works

Pilot Shell acts as a coordination layer that integrates with Claude Code and Codex. It uses a combination of hooks, stop guards, and independent reviews to prevent premature handoffs. It leverages a professional toolchain (including Semble, CodeGraph, and ast-grep) and browser automation for runtime verification.

It offers three ways of working: direct requests, native agent tools, and structured Pilot workflows (such as /spec, /build, /fix, and /prd) that create durable artifacts in the project's docs/ folder. A local Console allows users to review requirements, annotate diffs, and manage shared project knowledge.

Who it’s for

Software engineers who use AI coding agents (specifically Claude Code and Codex) and want to move beyond simple prompts to a rigorous, production-ready development workflow.

Highlights

  • Multi-Agent Support: Works as a unified system for both Claude Code and Codex.
  • Structured Workflows: Dedicated paths for specifications, building, fixing defects with TDD, and PRD creation.
  • Runtime Verification: Integrates tests, builds, and browser automation to turn claims into evidence.
  • Persistent Knowledge: Automatically captures and remembers findings across sessions and agents.
  • Visual Engineering: Integrates Open Claude Design and Impeccable for synchronized visual iteration.
  • Management Console: A local dashboard for reviewing progress, costs, and project memory.

Related

  • Project
    Arch1eSUN/Arcgentic

    A harness engineering layer for Codex and Claude Code that transforms ad-hoc AI prompting into a gated, role-based engineering workflow with audits and stop states.

  • Project
    affaan-m/ECC

    An agent harness operating system that provides AI coding agents with a structured engineering system, specialized skills, and a rigorous plan-test-verify workflow.

  • Dispatch
    Scaling AI Coding: How Claude Code Navigates Large Codebases

    An exploration of agentic search and the 'harness' architecture used by Claude Code to manage multi-million line monorepos and legacy systems without relying on traditional RAG indexing.

  • Project
    iamfakeguru/agent-md

    agent‑md is a lightweight framework that installs rule files, git‑hooks, and helper scripts to turn AI coding assistants (Claude Code, Codex, Cursor, Windsurf) into contract‑driven agents. It provides a single source‑of‑truth markdown file (AGENT.md) plus enforceable checks (type‑check, lint, tests, visual screenshots, state updates) that block the model until the required evidence is present.

  • Project
    Chachamaru127/claude-code-harness

    Claude Code Harness is a Go‑native plugin that orchestrates Claude‑based (or compatible) AI agents through a disciplined loop—plan, work, review, sync, release—while enforcing strong guardrails and human approvals at each step. It supports Claude Code, Codex CLI, Cursor, and Grok, generating specs, running AI‑driven implementations, independently reviewing changes, and packaging verified releases.