skillward · emploidai Marketplace
emploidai Marketplace
Add-onsAppletsPlugins
Search tools, teams, and capabilitiesPublish
MarketplacePluginsskillward
Plugin
Verified
Limited listing

skillward

by fangcunai · v0.0.10

SkillWard enables security review of AI Agent Skills and MCP projects before they are published or deployed, reducing the potential risks of Agent usage. Beyond static analysis and LLM evaluation, it executes suspicious Skills in isolated Docker sandboxes, replacing uncertain warnings with runtime evidence.

122 installsUpdated Jul 10, 2026
Documentation
Publisher information is incomplete

This community listing does not yet include every recommended support, privacy, pricing, and permission disclosure. Review the available package permissions before installing.

Capabilities

Tools

Available inside your emploidai workspace after installation.

Data sources

Available inside your emploidai workspace after installation.

Category

tool

Version

0.0.10fangcunai

Requirements

Maximum memory 256MB

Pricing

Not disclosed by publisher

Security & access

Review before installing

CompatibleRequires emploidai 1.0.0+

Permissions

  • Uses app capability
  • Uses endpoint capability
  • Uses model capability
  • Uses storage capability
  • Uses tool capability
  • Requires encrypted tool credentials

Dependencies

No additional dependencies

Resources

DocumentationPrivacy policy
emploidai Marketplace

Discover capabilities. Review access. Install inside your workspace.

DocumentationSecuritySupportPrivacyTerms

SkillWard

Author: Fangcun-AI Version: 0.0.10 Type: Tool plugin

SkillWard scans AI Agent Skills and MCP projects before they are published or deployed. This Dify tool plugin connects a Dify workflow to a running SkillWard Sentinel service and returns a structured security report.

This Marketplace release focuses on one stable workflow: provide a Skill or MCP source from Dify and scan it through the hosted SkillWard Sentinel backend.

Current Scope

Version 0.0.10 exposes the scan_skill_archive tool only.

The tool accepts exactly one source: an uploaded Skill/MCP .zip archive or one URL. The plugin forwards uploads as file with target_type=package, and forwards URLs as target with the selected target_type to the SkillWard Sentinel /api/scan/archive endpoint.

Roadmap

SkillWard is under active development. A major update is planned after this initial Marketplace release. The next major version is expected to improve the hosted scanning workflow and provide a more polished security report experience.

Setup

Configure the provider credential:

  • llm_api_key: OpenAI-compatible API key required by SkillWard Sentinel.
  • llm_base_url: OpenAI-compatible API base URL.
  • llm_model: Model name.
  • llm_api_version: Optional API version for providers that require one.

The hosted SkillWard Sentinel endpoint is built into the plugin and does not need to be configured in Dify.

Usage

Add the Scan Skill Source tool to a Dify Workflow, Chatflow, or Agent.

Provide exactly one scan source:

  • archive_file: A single uploaded Skill/MCP .zip archive. The backend automatically detects whether it is a Skill or MCP project.
  • target_url: A Git repository URL or remote MCP URL.
  • url_type: Required when target_url is provided. Use git_repo for Git repositories or remote_mcp for remote MCP URLs.

archive_file and target_url are mutually exclusive. Exactly one of them is required.

Expected scan options:

  • use_llm: Whether to enable SkillWard LLM safety evaluation.
  • use_runtime: Whether to enable Docker sandbox runtime verification.
  • enable_after_tool: Whether to enable post-runtime tool-chain analysis.
  • lang: Report language, en or zh.

Expected output:

  • summary: A short human-readable scan summary.
  • report: The full structured SkillWard scan report.
  • verdict: The SkillWard verdict, such as SAFE or UNSAFE.
  • skill_name: Name of the scanned Skill when available.
  • error: Error details when the scan request fails.

Hosted Backend

With the hosted SkillWard Sentinel backend, MCP security scanning, model evaluation, and Docker sandbox runtime checks run in the SkillWard API service. They do not run inside the Dify plugin process.

The Dify plugin only packages the request, sends it to SkillWard Sentinel, and returns the result to the workflow.

Self-Hosted Backend

If you are testing against a local SkillWard backend, start guardian-api first:

cd /path/to/SkillWard/guardian-api
python guardian_api.py --port 8899

Verify the service:

curl http://localhost:8899/api/health

Then configure the required LLM credentials in Dify.