Agent Skills · AI Agent

Llama Factory Wrap

@user_922b1001/llama-factory-wrap

Llama Factory Wrap is a tool skill related to Llama Factory, GitHub, and automation.

My MBS Self-Improvement Log

@user_3149ce24/mbs-ms

Lets AI agents log errors, corrections, and feature requests to Markdown and distill them into reusable project memory.

Wrap Cortex Multi-Tenant Prometheus Wrapper

@user_922b1001/wrap-cortex

Provides a wrapper for Cortex multi-tenant Prometheus for tool and automation scenarios.

Text Generation Inference Wrap

@user_922b1001/text-generation-inference-wrap

Wraps Hugging Face Text Generation Inference for use in agent tools and automated workflows.

Domain Payload Generator

@user_e02e04b8/domain-payload-generator

Collects fields step by step and generates domain payload data that can be written to JSON.

AI Love Companion

@user_b6f0de66/love-companion

A large-model skill pack for AI love-companion scenarios, installable as an Agent Skill.

TeddyMobile Vox Phone Notification

@user_caf52609/teddymobile-vox-phone-notification

A Claw phone-notification agent with dry-run parsing, trial outbound calls, and TeddyMobile live setup.

Less Token Prompt Compression

@user_676aba5c/ilang-less-token

Compress verbose summarization prompts into one-line structured instructions to save 40-65% tokens across LLM platforms.

WeChat Message Push and Auto-Reply

@user_0eb10066/wxskill

WeChat messaging capabilities for engineering integration, supporting group sending, entry push, and auto-reply with a unified JSON response.

Claude Code Dual-Agent Collaboration

@user_dc515faf/cc-wb-dual-agent

A safe advisory workflow where WorkBuddy executes and validates while Claude Code or other read-only advisors provide architecture, code review, and risk analysis.

Create Interesting Skills

@user_1e17f15e/create-interest-skills

Mine community ideas, assess feasibility and safety, then generate and deploy fun, practical OpenClaw skills.

Agent Self-Improvement Log

@user_28fe638b/zhao10

Log errors, corrections, and feature requests to Markdown, then promote durable rules to project memory files.