Build Enterprise Project

English · 中文

An evidence-driven, risk-scaled Codex skill for taking software projects from an idea or inherited repository through architecture, implementation, verification, release, production launch, operations, recovery, and continuous improvement.

一个以证据为依据、按风险分级的 Codex skill，引导软件项目从想法或既有仓库走到架构、实现、验证、发布、生产上线、运营、恢复与持续改进。

What it provides

• A progressive SKILL.md entry point for Codex.
• Risk-tiered delivery controls from T0 experiments through T3 high-assurance work.
• Lifecycle gates G0–G9 covering authority, product contracts, architecture, local reproducibility, vertical slices, quality, release, launch, and operations.
• An evidence-ledger model with read-only readiness evaluation and canonical digest tooling.
• Reference routing for data, security, reliability, frontend quality, AI/providers, metrics, incidents, release recovery, and platform-specific systems.
• Reusable templates for charters, ADRs, incident response, production readiness, and evidence.

This is a reasoning and verification aid, not a compliance certification or a substitute for product-specific engineering judgment, security review, legal advice, or human release authority.

When to use it

Use $build-enterprise-project when you are:

• starting a greenfield product or taking over an inherited repository;
• implementing a cross-cutting feature or productionizing an existing system;
• changing data, infrastructure, identity, domains, providers, queues, or release controls;
• auditing readiness, preparing a rehearsal, launching, operating, or recovering a system; or
• unblocking a local-to-production path while preserving evidence and recovery options.

Install as a Codex skill

Clone this repository and place the repository directory in your local Codex skills directory. Keep SKILL.md at the skill root together with references/, assets/, scripts/, and agents/.

git clone https://github.com/Hughhhhcoder/Build-Enterprise-Project.git

Then invoke the skill with $build-enterprise-project in Codex. The English SKILL.md is the canonical execution entry point; README.zh-CN.md is the Chinese reading guide.

Repository layout

Path	Purpose
SKILL.md	Canonical Codex instructions and frontmatter.
agents/openai.yaml	Display metadata and the default invocation prompt.
references/	Progressive lifecycle and domain-specific guidance.
assets/	Evidence, incident, charter, ADR, and readiness templates.
scripts/inventory_project.py	Sanitized, read-only repository inventory.
scripts/evaluate_readiness.py	Fail-closed milestone-readiness evaluator.
scripts/ledger_digest.py	Canonical, read-only ledger digest generator.
scripts/test_tools.py	Self-contained regression and structure test suite.
README.zh-CN.md	Chinese reading guide.

Validate locally

The helper scripts use only the Python standard library.

python3 -m py_compile scripts/*.py
python3 scripts/test_tools.py
python3 scripts/inventory_project.py /absolute/path/to/repository
python3 scripts/evaluate_readiness.py /absolute/path/to/evidence-ledger.json
python3 scripts/ledger_digest.py /absolute/path/to/evidence-ledger.json

inventory_project.py, evaluate_readiness.py, and ledger_digest.py are read-only with respect to their target inputs. Read and inspect their output before treating it as evidence.

Core operating principles

1. Establish authority and scope before changing code, data, infrastructure, external services, or production state.
2. Inspect the real repository and installed toolchain instead of assuming a framework, layout, database, or deployment platform.
3. Keep secrets out of source, fixtures, logs, evidence, and chat.
4. Treat unknown, stale, partial, and not run as unresolved rather than as passing evidence.
5. Prefer the smallest reversible end-to-end slice, with explicit invariants, observation, reconciliation, and recovery.
6. End material work with verified outcomes, remaining unknowns, risks, and the next safe action.

Topics / 主题

GitHub topics are kept in English for discoverability and follow GitHub's lowercase, hyphenated topic format. The Chinese glosses below make the classification explicit for bilingual readers.

codex-skill（Codex 技能） · ai-agent（AI Agent） · openai（OpenAI） · software-engineering（软件工程） · production-readiness（生产就绪） · software-architecture（软件架构） · devops（DevOps） · reliability-engineering（可靠性工程） · security-engineering（安全工程） · release-management（发布管理） · risk-management（风险管理） · developer-tools（开发者工具） · sre（站点可靠性工程） · open-source（开源）

Contributing and security

Please read CONTRIBUTING.md before opening a pull request. Report security vulnerabilities privately according to SECURITY.md, not in a public issue.

License

This project is released under the MIT License.