Skip to main content

Documentation Index

Fetch the complete documentation index at: https://hypha-docs.plinqx.app/llms.txt

Use this file to discover all available pages before exploring further.

Store Exposure determines which AI platforms (ChatGPT, Claude, Gemini) can discover and use your tools. Configure exposure per-tool to control visibility.

​
What is Store Exposure?

Store Exposure is a per-tool setting that controls:
  • Discovery: Whether AI platforms can see the tool in their tool lists
  • Access: Whether the platform can execute the tool
  • Risk Management: Configure risk levels and safety annotations per store

​
Supported Platforms

Plinqx supports three AI platforms:
PlatformDescription
ChatGPTOpenAI’s GPT Store and ChatGPT integrations
ClaudeAnthropic’s Claude Desktop and API integrations
Google GeminiGoogle’s Gemini AI platform

​
Configuring Store Exposure

​
Per-Tool Configuration

Configure exposure when creating or editing tools:
  • Custom Tools: Set in Step 4 (Response Mapping) of the wizard
  • Imported Tools: Toggle exposure indicators on the import group page
  • Flow Tools: Configure in Flow MCP Settings
  • Collection Tools: Set via Store Exposure Management page

​
Bulk Management

Use the Store Exposure Management page for bulk operations:
1

Navigate to Store Exposure

Click Store Exposure in the sidebar (under Tools).
2

Select Store

Choose the store tab (ChatGPT, Claude, or Gemini).
3

Toggle Exposure

Toggle tools on/off for the selected store.

​
Risk Levels

For each store exposure, you can set a risk level:
  • LOW: Safe operations, read-only or low-impact
  • MEDIUM: Moderate risk, may modify data
  • HIGH: High risk, destructive or sensitive operations
Use risk levels to help AI platforms understand tool safety and make better decisions.

​
Safety Annotations (Claude)

For Claude/Anthropic, you can add safety annotations:
  • Read-only: Tool only reads data, doesn’t modify
  • Destructive: Tool may perform destructive operations
  • Idempotent: Same call produces same result
  • Open World: Tool interacts with external entities

​
Best Practices

Start Conservative

Only expose tools that are ready for production use

Use Risk Levels

Set appropriate risk levels for each tool

Test First

Test tools in Playground before exposing to stores

Monitor Usage

Review tool usage per store to optimize exposure

​
Next Steps

Connect ChatGPT

Set up ChatGPT integration

Connect Claude

Set up Claude integration