Lazycat Skills
A skill pack for building apps on the LazyCat MicroServer platform: LPK packaging, routing, authentication and AI Pod.
- 61 ★ on the repository
- Last commit
Minimal access
Why this level
- The skills hold instructions and helper scripts for doc checking; by themselves they do not run platform code.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/lazycat-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skills hold instructions and helper scripts for doc checking; by themselves they do not run platform code.
This is third-party code. Review the repository files before installing.
What it does
The pack teaches an agent to build apps for LazyCat MicroServer, a Chinese home server platform. The `lazycat-lpk-builder` skill turns a Docker project or source into an `.lpk` package, and the entry skill `lazycat-developer-expert` routes a task to the right specialist. Separate skills cover dynamic deployment, routing by path prefix and host with TCP and UDP forwarding, OIDC sign-in with trusted identity headers, AI Pod resources and the Android WebView bridge. The skills that touch platform settings carry a procedure for checking against the official docs: the agent compares the spec snapshot with the source and reports differences.
Who it is for. App developers for LazyCat MicroServer who package Docker projects into LPK and configure routes and sign-in.
Good fit when
- A ready Docker project has to be packaged into `.lpk` with `package.yml` and `lzc-manifest.yml`.
- Traffic must be split by path prefixes with TCP or UDP forwarding.
- An app must accept OIDC sign-in and trusted platform headers.
- An AI Pod resource is being developed with separate compose sets for device models.
Not a fit when
- The LazyCat platform is not used: the skills are entirely about its package format.
- The old LPK V1 specification is needed: the main branch describes V2.
- The agent cannot reach the network while the exact spec version matters: the check falls back to the snapshot.
Example request
Package my Docker project for LazyCat: prepare `package.yml` and `lzc-manifest.yml` and check routing by the `/api` prefix.Limitations
The skills target LPK V2: lzcos 1.5.0 or newer and lzc-cli 2.0.0 or newer are required, and the V1 specification lives in a separate `v1` branch. The documentation in the pack is a snapshot from October 8, 2026 and may lag the official one. The author warns that the check procedure does not guarantee agent compliance and does not replace testing on a real device. The README is written in Chinese. No license file was found in the repository.
How to disable. Delete the `lazycat-*` folders from the project or agent skills directory. Updating needs a reinstall, because the pack does not update itself.
FAQ
How do I install the Lazycat skills?
Run `npx skills add whoamihappyhacking/lazycat-skills` in the project root. Agents with skill support discover them after installation.
What does lazycat-lpk-builder do?
It turns a Docker project or source code into an `.lpk` package for the platform, assembling the required descriptor files.
How do main and v1 branches differ?
The `main` branch describes LPK V2: lzcos 1.5.0 or newer, lzc-cli 2.0.0 or newer. The `v1` branch keeps the old LPK V1 specification.
How do the skills stay current?
Before editing settings the agent reads `references/spec-sync.md`, fetches the official text from the docs source repository and compares it with the snapshot. Without network it must declare it works from the snapshot.
Which agents fit?
The README names Cursor, Windsurf and Cline, but the skills work with any agent that supports Agent Skills.
Discussion
No questions or reviews yet.
Related
SkillsDocker Skills
docker-skills
Official Docker skills for agents: Dockerfile and builds, Compose, Sandboxes, Docker Agent and a skill that confirms irreversible actions
Plugin with one skill: the agent prepares a project, deploys a container to the Cloudish cloud and returns the live URL.
A small server and the skb CLI for private skill distribution: publish, version, install and roll back skills across agents
KubeShark skill: the agent finds the failure mode first, then writes manifests, Helm and Kustomize to official practice with a rollback plan
Official AMD skills teach an agent to set up ROCm, serve LLMs on Instinct and EPYC, profile GPUs and quantize models.
Cloudflare's official agent skills: Workers, Durable Objects, wrangler, Pages, storage and an MCP server
README in short
Installation is a single `npx skills add` command. After that you only ask the agent to package a project and it picks the right skill.