DOCSEARCHbyAlgolia
Submit Docs

Freshworks Developer Documentation

website·Indexed 19 days ago

https://developers.freshworks.com/docs/

Documentation for building custom apps on the Freshworks platform using the Freshworks Development Kit (FDK). Includes guides on app creation, testing, and deployment, as well as detailed App SDK references for versions v2.3 and v3.0. Covers serverless apps, scheduled and product events, Server Method Invocation (SMI), and AI agent skills.

Tokens
2.1M
Snippets
6.4K
Records
15.8K
Agent score
84%

What's inside Freshworks Developer Platform

  1. Overview of Freshworks Paid App benefits and infrastructure

    Developing paid apps for the Freshworks Marketplace provides developers with a managed ecosystem that handles the operational overhead of SaaS distribution. Key benefits include access to a global customer base (75,000+ customers in 120+ countries), recurring revenue streams, and a built-in distribution channel via the Marketplace. Freshworks manages the critical infrastructure, including payment processing, billing, SSL certificates, CDN, hosting, and automated update delivery, allowing developers to focus on feature development rather than DevOps.

  2. Overview of the Freshworks Chat Conversation Developer Toolkit

    v3.0
    The Freshworks developer platform provides a comprehensive toolkit for building apps within Chat Conversations. This toolkit includes AI capabilities (Freddy AI Copilot, Agentic Developer Toolkit), basic and advanced interfaces, a data store, and a front-end framework to streamline the development process.

  3. Overview of Freddy Copilot for Developers

    Freddy Copilot for Developers is a Generative AI-powered assistant integrated into Visual Studio Code for the Freshworks Developer Platform. It supports Node JS and React-based applications. The tool provides chat-based tutorials, rapid code generation, and one-click actions to help developers move from use-case definitions to code deployment more efficiently.

  4. Overview of Serverless Jobs for long-running tasks

    v3.0
    The Jobs feature allows serverless functions to run asynchronously for extended durations (up to 2 minutes), bypassing default execution time limits. This is ideal for resource-intensive operations such as bulk data imports, exports, or batch actions across multiple records. While the job runs in the background, the app receives status responses to track execution.

  6. Overview of Freshservice Serverless Jobs

    v2.3
    The Jobs feature allows serverless apps to execute time-consuming or resource-intensive tasks asynchronously. This is ideal for bulk actions like importing/exporting data or batch processing records. Jobs can run for up to 2 minutes, overcoming the default execution time limits of standard serverless functions. While a job runs in the background, the app receives a status response indicating the current state of execution.

  7. Overview of the Freshworks Support Agent Developer Toolkit

    v3.0
    The Freshworks developer platform provides a comprehensive toolkit for building apps. Key components include the Freshworks Agentic Developer Toolkit, Freddy AI Copilot for Developers, basic development tools, fundamental and advanced interfaces, a data store, and a front-end framework. In addition to these platform features, developers can utilize Public REST APIs to interact with product entities (such as tickets, contacts, and agents) and the CRAYONS component library for UI design.

  8. Overview of AI Actions for AI Agent Studio

    v3.0

    AI Actions apps allow developers to expose typed, callable operations to AI Agent Studio (the workspace for building Freddy AI-powered agents). These actions are used within agent workflows to perform tasks like updating records, resetting passwords, or executing custom business logic via the 'API action' block.

    Key architectural requirements:

    • Declaration: Each action must be declared in actions.json.
    • Implementation: The logic for each action must be implemented as a callback in server/server.js.
    • Naming: The action key in actions.json must exactly match the exported callback name in server/server.js to avoid fdk validate or runtime failures.
    • External APIs: If the action needs to call external APIs, request templates must be used and registered in manifest.json under modules.common.requests.
    • Deployment: AI Actions apps can currently only be published as Custom apps; publishing as Freshworks (public) apps is not supported.

  10. Overview of AI Actions for AI Agent Studio

    v3.0

    AI Actions allow developers to expose typed, callable operations to AI Agent Studio, which is the workspace for building Freddy AI-powered agents. These actions are used within agent workflows (via the API action block) to perform tasks such as resetting passwords, updating records, or executing custom business logic.

    Key architectural components include:

    • actions.json: Used to declare the input and output schemas for each action.
    • server/server.js: Contains the implementation logic (callbacks) for the actions.
    • Request Templates: Optional mechanism to connect actions to external APIs (must be registered in manifest.json > modules.common.requests).

    Crucially, the action key defined in actions.json must exactly match the exported callback name in server/server.js to avoid validation or runtime failures.

  11. Overview of Key-Value Storage in Freshworks App SDK

    v3.0

    The Key-Value (KV) store allows apps to persist and retrieve data on a per-account and per-app basis. Data stored by one app is not accessible by another app installed in the same account.

    Constraints and Limitations:

    • Rate Limit: 50 requests per minute per installed app (applies to set, get, and delete operations).
    • Cross-SKU Access: In production, modules deployed on different SKUs (product bundles) cannot access each other's KV store data, even if they are part of the same global app. Ensure app logic does not rely on cross-module data retrieval if modules are deployed to different SKUs.
    • Requirement: Requires CLI v4.1.2 or higher.