xsserandCodex
4a7793859b
Retain original skill rhetoric and the frozen flavor library while adding portable Claude Code, Codex and ChatGPT runtime contracts and packages. Correct hook event handling, scoped numeric checkpoints, flavor locks, plugin path resolution, advisory integrity checks and voluntary feedback. Add native-loading and ordered-evidence runners for cc0, OMP and Codex. Document paired model results, refusals, approval/rate-limit failures, explanatory factual errors and remaining evidence gaps without an all-model pass claim. Exclude private execution archives from public Git content. Bump all six PUA manifests including Pi, refresh three-language READMEs, and add release notes plus reproducible offline validation instructions. Resolve optional OMP source metadata relative to the current user's home. Validation: - 17 offline suites passed, including 13 OMP evidence regressions - Six manifest versions synchronized; all tested skill hashes unchanged - Three portable packages rebuilt reproducibly; original tone checks pass - Python, Bash and JSON syntax; public doc links; staged diff checks pass - 330 private archive files verified unchanged and excluded from staging Real-model results remain limited: this is not universal behavior acceptance. Co-authored-by: Codex <codex@openai.com>
2.5 KiB
构建与离线验收
1. 环境与范围
在包含完整 Git 历史的仓库根目录运行。语气保真测试读取固定上游提交
ac5026791845b730a18eb4ff07512a3b6f2f06f5；只有源码 ZIP 或缺少该提交的浅克隆不能完成这项检查。
需要 Git、Bash、Python 3.10+、jq 和 Python 的 PyYAML（YAML 解析库）。可在自己选择的虚拟环境中安装 PyYAML>=6,<7；测试不会自动安装依赖。以下命令不调用真实模型，不需要模型认证。
2. 一次完成构建与 17 组离线检查
set -euo pipefail
python3 scripts/build-model-compat.py </dev/null
for name in \
agent-governance integrity-guard issue-regressions microsoft-flavor \
no-telemetry platform-compat pua-loop-hook release-consistency \
trigger-regex windows-python-hooks yaml-frontmatter; do
bash "evals/test-${name}.sh" </dev/null
done
for name in cc0-fable codex-evidence feedback-runtime hook-runtime model-compat omp-evidence; do
python3 "evals/test-${name}.py" </dev/null
done
git diff --check
dist/ 由构建器生成且不提交：三个 ZIP 为独立技能包，manifest.json 记录实际成员和文件指纹。完整 Claude Code 插件的钩子和命令在仓库的 hooks/、commands/ 中，不在独立技能 ZIP 里。
其中 Windows 检查是本机模拟兼容回归，不等于在真实 Windows 上完成全流程安装。通过条件是所有命令正常退出、源文件与包一致、两次构建结果相同、原话术不变量保留；不把离线检查数量当作真实模型样本数。
3. 在线评估必须另外授权
run-trigger-test.sh、test-behavior.sh 会调用真实模型，不属于上面的离线列表。
run-cc0-fable.py、run-omp-pua.py、run-codex-pua.py 等运行器默认只预览，显式加 --run 才执行；具体参数先看各自 --help。cc0 是本地包装函数，不是本项目提供的通用命令。
按 原生加载验收方法 准备新的隔离目录、固定提示和外置可信清单，再分别检查身份、加载、功能、语气、执行顺序与权限。工具白名单不是操作系统沙箱，独立事后检查也不能倒算成模型自己运行过。
当前结果及限制见 模型验收矩阵。compat/evidence/ 是忽略的本地私有留档，不是公开仓库的缺失依赖；离线测试不依赖该目录。发布前审查待提交文件，不要用强制添加把会话、包装函数、凭据或原始输出带入 Git。