Membrane Application Skills

Membrane Application Skills are specialized instruction sets that teach AI agents how to interact with specific third-party applications (e.g., Gmail, Slack, Salesforce). Each skill provides the agent with knowledge of the app's structure (entities and resources), available actions, and guidance on when to use specific actions.

Key benefits include:

• Automated Authentication: Membrane manages OAuth, API keys, and token refreshes, so you don't need to provide credentials in your prompts.
• Agent Compatibility: Skills follow the Agent Skills specification, making them compatible with OpenClaw, Cursor, Claude Code, GitHub Copilot, Gemini CLI, and other supported agents.
• Direct Action Execution: Agents can translate natural language requests (e.g., "Send a message to #general on Slack") into specific API actions with correct parameters.

The Zendesk Sell skill allows you to manage data, records, and automate workflows within Zendesk Sell. It is used to interact with various CRM entities.

Note: The overview list provided in the documentation contains a vast range of conceptual terms; however, for practical integration, focus on core CRM entities such as Lead, Contact, Deal, Task, and Note.

The Lattice skill allows users to interact with data from the Lattice performance management platform. It is used to manage various organizational and HR-related entities.

Key capabilities include managing:

• People & Roles: Persons, Roles, Levels, Teams, Departments, Users, Employees, Admins.
• Performance & Feedback: Goals, Praise, Recognition, Feedback Requests, Review Cycles, Competency Assessments, Performance Reviews.
• Tasks & Activities: Tasks, Action Items, Projects, Meetings, Events, Schedule Items.
• Organizational Data: Organizations, Groups, Locations, Workspaces.
• Documents & Files: Documents, Templates, Files, Notes.

Note: This integration requires network access and a valid Membrane account (Free tier supported).

The Drata skill allows users to interact with Drata, a security and compliance automation platform. It enables management of various compliance-related entities including Controls, Standards, Objectives, Reports, Persons, Risks, and more.

Compatibility Requirements:

• Network access
• A valid Membrane account (Free tier is supported)

Use Case: Use this skill when you need to retrieve or manipulate data within a Drata environment to maintain security posture and compliance certifications (e.g., SOC 2, ISO 27001, HIPAA).

Follow these patterns when building with GorillaStack:

• Prefer Membrane actions over raw API calls: Pre-built actions include built-in authentication, pagination, and error handling, which is more efficient and secure.
• Discover before building: Always run membrane action list --intent=QUERY to check for existing actions before implementing custom API logic.
• Let Membrane handle credentials: Never request API keys or tokens from users. Use the membrane connect workflow so Membrane can manage the authentication lifecycle server-side.

To ensure security, efficiency, and reliability when integrating with Scrape-It.Cloud:

1. Prefer Membrane Actions: Use pre-built actions whenever possible. They include built-in authentication, pagination, and error handling, which reduces token usage and increases security.
2. Discover Before Building: Always run membrane action list --intent="QUERY" to check for existing actions before writing custom API calls or proxy requests.
3. Let Membrane Handle Credentials: Never request or store API keys or tokens locally. Use membrane connection to manage the authentication lifecycle server-side.

When building integrations with Mailercloud via Membrane, follow these principles:

1. Prefer Membrane Actions over Proxy Requests: Always use membrane action list to find existing actions first. Pre-built actions include built-in authentication, pagination, and error handling, which are more token-efficient and secure than raw API calls.
2. Discover Before Building: Use membrane action list --intent="QUERY" to see if your desired functionality is already available.
3. Let Membrane Handle Credentials: Never ask users for API keys or tokens directly. Use membrane connection ensure to let Membrane manage the full authentication lifecycle server-side.

When building integrations with Neetoinvoice, follow these principles:

1. Prefer Membrane Actions: Always use membrane action list to find existing actions before writing custom API calls. Pre-built actions handle pagination, field mapping, and error handling more efficiently and use fewer tokens.
2. Discover Before Building: Use membrane action list --intent="QUERY" to explore capabilities.
3. Let Membrane Handle Credentials: Never ask users for API keys or tokens directly. Use membrane connection ensure to let Membrane manage the full authentication lifecycle server-side.

To ensure security, efficiency, and reliability when working with ABRA FlexiBee:

1. Prefer Membrane Actions: Always use membrane action list to find pre-built actions before writing custom API calls. Actions include built-in authentication, pagination, and error handling, which saves tokens and is more secure.
2. Discover Before Building: Use membrane action list --intent=QUERY to find existing capabilities.
3. Let Membrane Handle Credentials: Never ask users for API keys or tokens directly. Use membrane connection to manage the full authentication lifecycle server-side without storing local secrets.

To ensure efficient and secure integrations:

1. Prefer Membrane Actions: Always use pre-built Membrane actions instead of raw API calls when possible. They include built-in authentication, pagination, and error handling, which reduces token usage and increases security.
2. Discover Before Building: Use membrane action list --intent="QUERY" to find existing actions before attempting to write custom proxy requests.
3. Let Membrane Handle Credentials: Never request API keys or tokens from the user. Use membrane connection to manage the authentication lifecycle server-side without exposing local secrets.

When building integrations with ScrapingAnt via Membrane, follow these principles:

1. Prefer Membrane Actions: Use pre-built actions whenever possible. They include built-in authentication, pagination, and error handling, which is more token-efficient and secure than raw API calls.
2. Discover Before Building: Always run membrane action list --intent="QUERY" to see if an existing action can satisfy your requirement before writing custom proxy requests.
3. Let Membrane Handle Credentials: Never request API keys or tokens from the user. Instead, create a connection; Membrane manages the entire authentication lifecycle server-side.