rdk-model-zoo-release
Skill404 starsChanged 30 days agoApache-2.0View on GitHub
Use when preparing or checking RDK Model Zoo model releases, Skills Pack releases, manifests, Hub registration migration, tags, or release isolation. 触发词:模型发版、Skills 发版、Hub 接入、Tag 兼容。Do not use for routine sample development, OE Pack releases, or publishing without explicit authorization.
1 other project uses a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- D-Robotics/rdk-skillsedited, first seen Oct 9, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 1
- D-Robotics/rdk-skills★ 594% same
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository6 other files it gives its agents.
- rdk-model-zoo-developSkill
- rdk-model-zoo-integrateSkill
- rdk-model-zoo-repoSkill
- rdk-model-zoo-reviewSkill
- rdk-model-zooSkill
- rdk-model-zoo-validateSkill
The file
Files
RDK Model Zoo Release
Purpose
准备可追溯、相互隔离的模型与 Skills 发布。默认只做检查和发布计划;本 Skill 不提供未经审批的自动发版权限。
When to use
模式 model-release:目标平台模型版本、Manifest、Benchmark 和目录发布。
模式 skills-release:本仓库 skills/ 的独立版本、结构/行为验收和正式 Release。
模式 hub-onboarding:组件注册、同名入口维护源迁移和安装/升级验收。
请求未指明“模型”还是“Skills”且无法从文件范围判断时,先明确发布对象,不默认修改根 VERSION。
Instructions
- 读取 context-policy.md、目标仓库发版规范和 release-contract.md。
main是交付默认分支,develop是统一集成/开发线,保留的rdk_x5/rdk_s是历史分板卡交付线;具体检出的 ref 与默认分支仍以目标检出当天的 Git 状态为准,尚未执行的切换不得在发布材料中写成已完成。按目标 ref 识别 X5、S100/S100P/S600、X3 或 legacy,并核对该 ref 的 Manifest(有清单时)。记录发布类型、候选提交、版本、原 Tag/Latest/Manifest 状态和权限;用户指定的对象冲突时不切换分支。 model-release只修改该平台版本线的 VERSION、CHANGELOG、适用布局的 Manifest 与发布说明;已发布 Tag 不变。统一集成线上各平台版本线位于docs/release/{x5,s}/VERSION,清单与基准数据位于docs/release/{x5,s}/models.yaml与benchmarks.yaml;历史 ref 按其实际布局。遵循当前 S/X3 固定来源与 X5 目录发布流程,不把静态清单构建写成全量板测。skills-release读取 skills/VERSION、技能自身版本与 skills/CHANGELOG;严格区分 Pack 版本和单 Skill 版本。Tag 命名各归其位:裸 vMAJOR.MINOR.PATCH 只表示 Skills Pack;统一仓库发布用 model-zoo-vMAJOR.MINOR.PATCH 并持有 Latest;web-v*是网站部署 tag,不代表 Pack 或模型版本;历史模型 x5-v*/s-v*/x3-v* 保持原样,不改名、不删除。核对技能自包含、引用、行为评测和许可/来源。- 验证发布隔离:Skills 不改模型 Manifest/根 VERSION;非模型事件不进入 pages 并发组;Skills Release 正式、非草稿、非预发布,显式 latest=false,Latest 归统一仓库 model-zoo-v* 发布。附注 Tag 指向批准的真实提交;不能把旧模型 Tag 重命名为 Skills Tag。
hub-onboarding核对组件 ref/源路径/catalog_dir 唯一。先发布真实 Skills 版本,再在同一个 Hub PR 中移除 Device 的 rdk-model-zoo 条目并新增 Model Zoo 组件。注册变更、镜像和生成物一起验证,不能依赖每小时同步会自动完成迁移。- 默认输出 release-checklist.md,列命令、审批点和失败回退。仓库级规则、App/Environment、发布通知和标签配置只读取;没有权限不能声称已配置。
- 只有用户明确授权了准确对象、提交、版本和副作用,才进入实际 Tag/Release/Hub 写入流程。使用只增不改的附注 Tag,创建 Release 需 verify-tag;Skills 必须 latest=false。发布前重验候选与审批后状态。
- 半发布失败保留 Tag;核对精确提交和原批准说明后才恢复缺失 Release,不删除重打。不移动旧 Tag 回滚;已发布缺陷用新补丁版本,Hub 回指旧合法 ref 需显式批准且不能伪装更高版本。
Output
发布对象与候选、检查证据、未满足项、需审批动作、回退和发布后核对项。没有执行就写“发布准备完成/未发布”,不能输出不存在的 Release URL。
Safety
禁止强推/移动/删除已发布 Tag;不将 GitHub 写权限视为批准。不开启自动合并、不绕过保护、不共享 App 密钥。模型和 Skills 发布/通知并发及 Latest 相互隔离;历史内容保留。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

