OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Infrastructure

Skillnet

Skillnet is open infrastructure for discovering, creating, evaluating, analyzing and routing reusable AI agent skills. It combines a public skill library with Python SDK and CLI tools for local skill workflows, plus a browser interface for inspecting skill files and saved analysis.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Skillnet — OpenViST project profile in Agent Infrastructure.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use the website to discover and review existing skills. Use the Python SDK or CLI when working with local skill folders, generating or assessing skills, analyzing relationships, or routing tasks. Use the local browser interface when you want to inspect skill files or view analysis already generated elsewhere.

Who it is for

  • Agent developers looking for reusable skills in a public library
  • Teams organizing and analyzing local skill collections
  • Builders who need to evaluate skills or select capabilities for tasks

Documented uses

  • Search public skills by keyword or semantic intent and filter by category
  • Review skill descriptions, evaluation ratings and GitHub sources before reuse
  • Download skill folders from GitHub into local agent workspaces
  • Create structured skills from repositories, documents, prompts or execution traces
  • Assess skills for safety, completeness, executability, maintainability and cost awareness
Before you choose

Requirements & limits

The public search API and public skill downloads require no authentication. The repository is public and identifies an MIT license. Python SDK and CLI operations that create, evaluate or analyze skills require an OpenAI-compatible Chat Completions endpoint; analysis and routing also need an embedding endpoint, and routing requires a Claude or Codex Agent SDK configuration.

  • The public library supports keyword and semantic-intent search, category filters, skill descriptions, evaluation ratings, GitHub sources and download links.
  • Creation, evaluation, analysis and routing have different endpoint and credential requirements; public search and downloads do not require an API key.
  • The local browser interface displays existing files and saved analysis rather than generating analysis.
  • The project requires Python 3.10 or newer for its SDK workflow.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Browse local skills with Skillnet's browser interface

Open a local skill folder and inspect its files or previously generated relationship analysis.

Before you start

  • Python 3.10 or newer
  • The skillnet-ai package with its UI extra, version 0.1.2 or newer
  • An absolute path to a local skills directory

Steps

  1. Install the UI package with the command below.

    Documented command
    pip install "skillnet-ai[ui]>=0.1.2"

  2. Launch it with skillnet ui --skills-dir "/absolute/path/to/skills".

  3. Open http://127.0.0.1:8765.

  4. Choose Current skill files or Analysis results; analysis results can also be loaded from a graph.json generated by skillnet analyze.

Expected outcome

A local browser view of skill files or saved relationships, with source evidence available for relationship inspection.

Guide limitations

  • The interface starts with an empty personal library.
  • It displays analysis results but does not generate them; generate analysis separately with the SDK or CLI.
  • The documented packaged interface does not require Node.js or model keys just to browse.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
zjunlp/SkillNet · Official homepage title · retained in its source language
How we review sourcesSuggest a correction