AGENTS.md - SteadyAgent Codex Guide

This file gives Codex a compact project map for contributing to SteadyAgent.

Project Intent

SteadyAgent is a local-first harness for AI coding agents. It packages concise instructions, progressive rules, validation scripts, lifecycle hook examples, and release evidence so agent-assisted development is easier to scope, verify, review, and recover.

Working Rules

• Keep changes scoped to the requested workflow, tool, template, or documentation surface.
• Read the nearest project guide, relevant docs, and directly touched scripts before editing.
• Prefer deterministic scripts, tests, and hooks over longer instructions when behavior must be enforced.
• Preserve the public release surface: no private paths, local credentials, unpublished plans, or maintainer-only notes.
• Keep English as the default public surface; update the paired Chinese docs when user-facing behavior changes.
• Use rg or rg --files for repository search.

Validation

Run the smallest relevant gate before committing:

powershell.exe -NoProfile -ExecutionPolicy Bypass -File .\tools\validate-release-readiness.ps1

Focused checks:

powershell.exe -NoProfile -ExecutionPolicy Bypass -File .\tools\validate-phase3.ps1
powershell.exe -NoProfile -ExecutionPolicy Bypass -File .\tools\validate-runtime-slice.ps1
powershell.exe -NoProfile -ExecutionPolicy Bypass -File .\tools\test-agent-hooks.ps1

Use -AllowDirty with validate-release-readiness.ps1 only while reviewing an intentional work-in-progress tree.

Release Discipline

• Do not publish, retag, or rewrite remote history without explicit maintainer approval.
• Treat v1.0.0 and later releases as audited artifacts: clean history, reproducible validation, and clear release notes.
• Use docs/github-publication-runbook.md for public release steps.