Public case repository for building a reusable AI/Agent-assisted development environment on cloud servers.
This repository is organized as a set of independent, reusable solution modules.
- Repository overview:
docs/overview.md - Current roadmap:
docs/roadmap.md - Screenshot publishing guidance:
docs/screenshot-guidelines.md - First complete solution:
solutions/file-sharing/
The current case modules are validated against Linux cloud servers, including Ubuntu and Alibaba Cloud Linux.
Most operational examples assume Linux server conventions such as systemd, Docker Engine / Docker Compose, Caddy on a public host, SSH access, and shell commands like ss, journalctl, and curl.
Future modules may cover macOS or Windows workflows, but those platforms should be documented as separate variants instead of silently reusing Linux-only commands.
This repository is intentionally starting small.
- First published module:
solutions/file-sharing - Newly added service deployment module:
solutions/sub2api-deployment - Current focus: practical, reproducible, public-safe solution examples
- Future direction: expand into a broader set of cloud-server AI/Agent environment patterns
solutions/file-sharing- sync local cloud-server text files to real Feishu/Lark Drive folders and online docs, with message notifications.solutions/sub2api-deployment- deploysub2apion a fresh cloud server with Docker Compose, low-memory guardrails, and recovery-oriented notes.solutions/web-access-routing- publish local AI tools and admin panels through subdomain-first Caddy reverse proxy patterns.
- Cloud-server AI/Agent workflows are useful, but often too tied to one machine to share cleanly.
- Rebuilding the same environment on another server is expensive if the setup only lives in ad hoc notes.
- Public examples are valuable, but only if they are stripped of secrets, local state, and operational identifiers.
This repository is meant to turn real working setups into reusable, public-safe case modules.
- Keep each solution independently understandable and reusable.
- Prefer simple operational recovery on a fresh cloud server.
- Separate public-safe templates from machine-local secrets and state.
- Preserve enough documentation so future AI/Agent sessions do not need to rediscover the setup.
- Make each workflow friendly to both human operators and AI agents: document direct commands, preview steps, confirmation points, and any skill installation or runtime restart requirements.
This repository is intended to grow into a broader collection of cloud-server AI/Agent environment patterns, such as:
- reverse proxy and web access patterns
- remote coding and agent entrypoints
- chat-driven operations
- environment bootstrap and disaster recovery
This repository intentionally excludes:
- real app secrets
- real access tokens
- real local state files
- machine-specific private configuration
Use the example configuration files as templates and create your own local copies.
This repository is released under the MIT License.
docs/ repository-level overview and guidelines
solutions/file-sharing/ sync server text files to Feishu/Lark docs
solutions/sub2api-deployment/ deploy sub2api on a fresh cloud server
solutions/web-access-routing/ subdomain-first reverse proxy patterns for local web apps
- English repository entry:
README.md - Chinese repository entry:
README.zh-CN.md - English file-sharing module entry:
solutions/file-sharing/README.md - Chinese file-sharing module entry:
solutions/file-sharing/README.zh-CN.md - English sub2api deployment module entry:
solutions/sub2api-deployment/README.md - Chinese sub2api deployment module entry:
solutions/sub2api-deployment/README.zh-CN.md - English web-access-routing module entry:
solutions/web-access-routing/README.md - Chinese web-access-routing module entry:
solutions/web-access-routing/README.zh-CN.md - Screenshot publishing guidance:
docs/screenshot-guidelines.md