Building on its previous launch, ImageKit introduces three specialized MCP servers, granular access controls, and ready-to-use Agent Skills to enable secure, standards-based AI workflows across digital asset management and media infrastructure.

ImageKit, the complete media lifecycle platform for websites and apps, today announced a major expansion of its Model Context Protocol (MCP) offering, introducing enhanced governance, granular access controls, broader operational capabilities, and a library of Agent Skills to support enterprise AI workflows.

Building on its initial MCP server launch last year, ImageKit has evolved its offering into three specialized, fully hosted MCP servers:

1. DAM MCP server: Enables AI assistants to search, organize, manage, share, and govern digital assets through natural-language instructions.
2. Admin MCP server: Enables authorized users to manage media delivery infrastructure, account configurations, and usage analytics through AI assistants.
3. DevTools MCP server: Provides AI coding assistants with access to current ImageKit documentation and media transformation URL generation, without requiring authentication or an ImageKit account connection.

ImageKit's DAM MCP server is now available in the Claude Directory, while all three MCP servers—DAM, Admin, and DevTools—are listed in the official MCP Registry. Organizations using other MCP-compatible clients, including Cursor, VS Code Copilot, Windsurf, and Codex, can connect using ImageKit's integration documentation.

As enterprises increasingly incorporate AI assistants into development, marketing, and content operations, secure and governed access to business-critical systems has become essential. Organizations need AI agents that can execute complex workflows while operating within established security policies, access permissions, and operational boundaries.

ImageKit's expanded MCP offering addresses these requirements through a standards-based architecture that combines granular authorization, broader operational capabilities, and purpose-built guidance for AI assistants.

"Our first MCP server made ImageKit's media capabilities accessible inside AI assistants. As organizations move from experimenting with AI agents to actually integrating them into everyday operations, secure and governed access becomes increasingly important," said Manu Chaudhary, co-founder and CTO of ImageKit. "With this expansion to over 70 DAM tools and Agent Skills, an AI assistant can now search, organize, and publish assets end to end, while working within the same permissions as the person who connected it, bringing the intelligence, control, flexibility, and reliability enterprise operations require."

Strengthening Enterprise Governance with Granular Access Controls

A key advancement in ImageKit's expanded MCP offering is a comprehensive authorization framework designed around the principle of least privilege.

Rather than requiring broad, account-wide access, ImageKit enables organizations to control precisely which resources an AI assistant can access and what actions it can perform.

Key governance and security capabilities include:

• Granular, scope-based permissions: Independently authorize view, create, update, or delete operations for specific resources, ensuring assistants receive only the access required for their tasks.
• Role-based access enforcement: MCP operations respect existing ImageKit user roles and DAM permissions, preventing assistants from exceeding established access privileges.
• Secure authentication and revocable access: OAuth-based authentication protects account credentials, while users can review and revoke connected applications or approve additional permissions as needed.
• Separation of operational privileges: DAM and Admin operate through independent connections with distinct permissions, keeping media management separate from sensitive infrastructure administration.
• Safeguards for destructive operations: Actions such as deleting assets or overwriting content are explicitly identified, enabling compatible MCP clients to request confirmation before execution.

These controls enable enterprises to introduce AI-powered automation while preserving existing governance policies and reducing the risks associated with unrestricted agent permissions.

Expanded Capabilities for AI-Powered Media Workflows

ImageKit's updated MCP architecture enables a broader range of media and infrastructure operations through natural-language instructions.

Key capabilities include:

• Advanced Agentic DAM capabilities: Access more than 70 DAM tools for searching media libraries, managing metadata and taxonomies, organizing files and folders, updating assets in bulk, controlling access, and managing publishing workflows.
• Interactive media workflows: Browse visual search results, preview assets, and upload files through interactive interfaces in compatible MCP clients, with uploads handled directly by ImageKit rather than passing file contents through AI conversations.
• Infrastructure and account administration: Manage storage origins, configure URL endpoints, and retrieve account usage and analytics through AI-assisted workflows.
• Documentation-backed developer assistance: Search current ImageKit documentation, API references, and SDK examples, and generate valid transformation URLs for resizing, cropping, background removal, generative fill, upscaling, overlays, and other media operations.

Introducing Agent Skills for More Reliable AI Workflows

Alongside its expanded MCP infrastructure, ImageKit has introduced eight ready-to-use Agent Skills designed to help AI assistants execute ImageKit-specific workflows more consistently and accurately.

While MCP servers provide access to ImageKit's tools and capabilities, Agent Skills provide task-specific instructions that guide assistants on how to use those capabilities effectively.

The available skills cover MCP server selection, advanced asset search, secure file uploads, asset access control, AI-powered metadata operations, documentation search, transformation URL generation, and SDK and integration selection.

For example, an AI assistant can use these skills to discover an organization's custom metadata structure before searching for assets, generate accurate transformation URLs using documented parameters, or select the appropriate SDK for a development environment.

By providing structured guidance for common workflows, Agent Skills help reduce errors associated with outdated API knowledge, incorrect search queries, and unsuitable implementation approaches.

The skills are available through ImageKit's public repository and support popular AI development environments, including Cursor, VS Code Copilot, Windsurf, Claude Code, and Codex.

Built on Open Standards for Enterprise Interoperability

ImageKit's expanded offering is built on the Model Context Protocol, an open standard for connecting AI applications with external tools and data.

The servers are fully hosted and managed by ImageKit, eliminating the need for organizations to deploy or maintain separate MCP infrastructure. Integration with existing ImageKit permissions allows teams to adopt AI-assisted workflows without introducing a parallel access-control system.

ImageKit also provides AI-accessible documentation in Markdown format, along with documentation indexes designed for language models, helping assistants retrieve current technical information rather than relying solely on potentially outdated training data.

Together, these enhancements advance ImageKit's broader vision for agentic digital asset management, extending intelligent media operations beyond its native platform into the AI assistants and development environments organizations already use.

Availability

ImageKit's expanded MCP servers and Agent Skills are available now. Users can follow the Build with AI documentation to connect them to supported AI assistants and install the available skills.

About ImageKit

ImageKit is a media lifecycle platform for websites and apps: one system for managing assets, producing on-brand creative, and delivering images and video to production, without an export or a handoff in between.

It works with the storage you already have, including S3, GCS, and your own server, and with a built-in Agentic DAM for the assets you want findable, governed, and operable by agents, tagged by LLMs in your business's own vocabulary rather than generic labels. Creative teams generate on-brand variants from templates, with GenAI where they need it. Developers get 50+ real-time transformations and adaptive streaming that works with any player. And through ImageKit's MCP servers, an agent can do the work end to end, from finding the approved assets and resizing them for every screen to publishing them live, under the same permissions as the person who asked.

300,000+ developers and 3,000 businesses in 85 countries run on ImageKit, with enterprise-grade reliability and without enterprise pricing.

Learn more at imagekit.io.

View source version on businesswire.com: https://www.businesswire.com/news/home/20261015865599/en/

With this expansion to over 70 DAM tools and Agent Skills, an AI assistant can now search, organize, and publish assets end to end, while working within the same permissions as the person who connected it.