x5-router
Skill1 starsChanged 3 months agoApache-2.0View on GitHub
路由 X5 环境、OE Mapper PTQ、Plugin QAT、Runtime、板端 Python 和诊断请求;当目标芯片明确为 X5 或请求包含 bayes-e、hb_mapper、X5 .bin、March.BAYES_E 时使用。只选择一个主 Skill 并生成 route.json;不执行 HAT、X3 或 S 系列工作流。
In this repository59 other files it gives its agents.
- bsp-bootloader-buildSkill
- bsp-deb-buildSkill
- bsp-env-setupSkill
- bsp-image-buildSkill
- bsp-kernel-buildSkill
- bsp-rootfs-customSkill
- bsp-source-syncSkill
- bsp-s-seriesSkill
- hmct-workflowSkill
- horizon-routerSkill
- ucpSkill
- x5-accuracy-diagnosticsSkill
- x5-board-monitorSkill
- x5-bpu-python-apiSkill
- x5-calibration-data-prepareSkill
- x5-consistency-diagnosticsSkill
- x5-environment-installSkill
- x5-environment-probeSkill
- x5-environment-setupSkill
- x5-model-diagnosticsSkill
- x5-model-preflightSkill
- x5-performance-diagnosticsSkill
- x5-ptq-compileSkill
- x5-ptq-config-authoringSkill
- x5-ptq-deploySkill
- x5-qat-adaptationSkill
- x5-qat-compileSkill
- x5-qat-deploySkill
- x5-qat-trainingSkill
- x5-runtime-cpp-inferSkill
- x5-runtime-deploySkill
- x5-runtime-perf-evalSkill
- rdk-accessoriesSkill
- rdk-board-delegateSkill
- rdk-board-knowledgeSkill
- rdk-camera-setupSkill
- rdk-command-manualSkill
- rdk-diagnosticSkill
- rdk-docs-referenceSkill
- rdk-ecosystemSkill
- rdk-embodied-lerobotSkill
- rdk-gpio-40pinSkill
- rdk-hardwareSkill
- rdk-headless-modeSkill
- rdk-llm-deploymentSkill
- rdk-log-forensicsSkill
- rdk-memory-auditSkill
- rdk-model-benchmarkSkill
- rdk-model-deploySkill
- rdk-model-zooSkill
- rdk-multimediaSkill
- rdk-network-remoteSkill
- rdk-peripheral-cookbookSkill
- rdk-source-mapSkill
- rdk-system-configSkill
- rdk-system-maintainSkill
- rdk-tros-setupSkill
- rdk-vision-pipelineSkill
- d-robotics-pack-installerSkill
The file
Files
X5 路由
目标与边界
- 读取
.drobotics/platforms/x5/skill-index.json,以输入、产物、环境和风险选择一个主 Skill。 - 用户只说“X 系列”时先确认 X5/X3;X3 和其他 X 芯片在专属 Pack 可用前返回
blocked。 - HAT、HAT config、Trainer、Model Zoo 和
tools/compile_perf.py不在范围内。 - 不调用
s-*、HBDK4、HMCT、UCP、nash-*或March.BAYES。
输入合同
- 用户结果目标、目标芯片、模型/代码路径和期望产物。
- 已知模型格式、输入输出合同、板端信息和失败收据。
- 是否允许安装、训练、生成模型、上传或覆盖文件。
前置检查
- 用芯片型号、模型后缀和工具名确认平台;
.bin不能单独证明 X5,仍需hb_model_info或来源证据。 - 区分 PTQ 与 QAT:ONNX/Caffe 浮点模型走 PTQ;可训练 PyTorch + Plugin 走 QAT。
- 区分 Runtime:
.binC/C++/命令行走 Runtime;板端HB_HBMRuntimePython 走 Python API。 - 环境未知时先选
x5-environment-probe,不要直接选安装或执行 Skill。
执行步骤
| 目标 | 主 Skill | 典型后续交接 |
| 检查/准备环境 | x5-environment-setup | probe → install(如获确认) |
ONNX/Caffe 到 X5 .bin | x5-ptq-deploy | preflight → data → config → compile → Runtime |
| Plugin calibration/QAT | x5-qat-deploy | adaptation → training → compile |
.bin C/C++/上板/性能 | x5-runtime-deploy | C++ → perf → monitor |
| X5 板端 Python | x5-bpu-python-api | environment/diagnostics |
| 已有失败证据 | x5-model-diagnostics | accuracy/consistency/performance |
将候选、拒绝理由、主 Skill 和有序 handoff 写入运行目录的 route.json。多个关键词命中时也只保留一个主 Skill。
产物与完成标准
route.json符合 Pack 的 route schema,且selected_skill在 X5 注册表中。- 明确说明拒绝 S、X3、HAT 或格式不兼容候选的理由。
- 环境、输入或授权不足时只路由到预检/探测,不生成执行成功结论。
风险与确认
路由本身为低风险。不得把用户对“分析”的同意解释为对安装、训练、覆盖、上传或停止进程的授权。
失败与交接
- 芯片不明:只问一个最小问题“目标板卡是 X5、X3 还是其他型号?”
- HAT 请求:说明当前 Pack 明确排除 HAT,不迁移配置。
- 无合法候选:保存拒绝理由并以
blocked结束。
按需参考
.drobotics/platforms/x5/policies/compatibility.md.drobotics/platforms/x5/policies/risk-policy.md.drobotics/platforms/x5/references/run-contract.md
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.

