skills.rest
kfallahkfallahCommunity·5 Agent Skills Included

CLaaS

Continual model learning from live user feedback

Distills user feedback into model weights so language models keep improving after every conversation. Runs local GPU, Tinker-hosted, or Modal training stacks with Docker, dashboards, and evaluation harnesses included. Removes manual retraining pipelines by turning chat feedback into automatic LoRA updates.
59|6
↗
npx skills add kfallah/CLaaS --all -g -y
Available:

Gives your AI agent the project's debugging philosophy, testing commands, deployment steps, and architecture rules so it can safely modify and operate the CLaaS training stack.

All Skills in This Repository (5)

Pure Emerald Level Indicators
📦 In Repo
kfallahkfallah

pr-feedback

Categorize and address GitHub pull request reviewer feedback with code fixes or explanations.

Community
Advanced
📦 In Repo
kfallahkfallah

clear-tinker-storage

Delete all training and sampler checkpoints from a Tinker account.

Community
Intermediate
📦 In Repo
kfallahkfallah

setup-local

Set up the CLaaS local environment with vLLM, API, and Telegram gateway.

Community
Advanced
📦 In Repo
kfallahkfallah

setup-modal

Deploy the CLaaS Modal serverless distillation API with DistillWorker and FastAPI.

Community
Intermediate
📦 In Repo
kfallahkfallah

setup-tinker

Deploy the CLaaS Tinker stack with Docker Compose for GPU-free operation.

Community
Intermediate

Frequently Asked Questions

FAQPage Schema
How to install CLaaS?▼

Run `npx skills add kfallah/CLaaS --all -g -y` in your terminal to install all skills in this suite globally.

What does CLaaS do?▼

CLaaS continually trains a language model from live user feedback by distilling corrections into LoRA weight updates, so the model retains what it learns between conversations.

Do I need a GPU to run CLaaS?▼

No. The Tinker engine runs hosted training with just an API key, while the local engine requires an NVIDIA GPU with at least 24 GB of VRAM.

How do I deploy the CLaaS training stack?▼

Use the setup-local, setup-tinker, or setup-modal skills, which configure Docker, environment variables, and inference services automatically from your plain-English request.

Can CLaaS measure whether training actually helps?▼

Yes. The built-in eval harness runs automated feedback loops and tracks metrics like preference alignment and collapse detection through web dashboards.

Related Repositories in Software Engineering

View All in Software Engineering→
openclawopenclaw
388.4k

openclaw

Run a personal AI assistant across your devices and chat apps

Official📦 120 Skills
obraobra
280.0k

superpowers

Gives coding agents a disciplined workflow from idea to merged code

Community📦 15 Skills
facebookfacebook
248.4k

react

AI agent skills for building, testing, and porting React core

Official📦 14 Skills