Can one plugin change how an AI agent treats every package it touches? That question got a real answer on June 10, 2026, when JFrog teamed up with Anthropic to launch a new plugin built for Claude Code. The plugin went live immediately, open to every Claude Code user. Until now, most AI agents have written code fast but have had little insight into where the building blocks of that code came from. This plugin changes that picture. It connects the Claude AI coding assistant straight to JFrog’s security scanning, package curation, and artefact tools, right inside the workflow developers already use daily. Now, agents can check, trace, and govern that code too. This blog breaks down what the plugin does, why it matters, and how to start using it.
Claude AI Coding Assistant: What JFrog’s Claude Code Plugin Actually Does
At its core, the JFrog plugin extends the power of the Claude AI coding assistant by connecting it directly to the JFrog Platform. Instead of working only inside the code editor, the assistant now gains structured access to real development and DevOps systems. This means it can interact with live project data, security layers, and build artifacts in a controlled and meaningful way.
The integration is built across three important layers that define how the system operates in real environments:
- Platform Skills for handling daily development and operational tasks across repositories and builds
- MCP Tools that provide standardised and secure access to critical system data and governance rules
- Agent-native plugins designed for tools like Claude Code, Cursor, and VS Code Copilot for a seamless developer experience
Together, these layers allow the Claude AI coding assistant to move beyond simple code suggestions. Developers can now interact with it using natural language commands to perform advanced tasks without switching tools.
For example, a developer can ask the Claude AI coding assistant to scan a dependency for vulnerabilities, check a specific CVE, or retrieve build details directly from Artifactory. All of this happens without opening separate dashboards or switching platforms.
As a result, the Claude AI coding assistant becomes more than just a coding helper. It evolves into a governed, context-aware teammate that understands both code and system-level operations.
Why Supply Chain Governance Matters for Claude Code Plugin Agents Today
Speed without checks is risky, and governance cannot wait. AI agents now move at a pace human reviewers cannot always match, which means every dependency they touch needs a safety net built into the workflow itself, not bolted on afterwards. A few numbers make the urgency clear:
- JFrog’s platform now manages more than 18 billion artifacts, up 136% from last year
- JFrog notes that AI agents, including the Claude AI coding assistant, often act without enough context about what they pull in
- That gap can let unsafe packages slip into production unnoticed, sometimes for weeks before anyone catches the issue
The case for governed AI coding agents for developers has never been stronger, especially as more teams hand over routine coding work to autonomous tools. A few of the biggest risks this plugin tackles include:
- Unverified packages are entering a build unchecked
- Vulnerable dependencies going unnoticed until late in the cycle, often after release
- Agents pulling code from public registries with no curation layer in place
- Limited visibility into what an autonomous agent actually changed across a codebase
- Compliance teams spend days producing audit-proof for a single release
Left unmanaged, these gaps compound quickly across large engineering organisations.
Core Features Inside the JFrog’s Claude Code Plugin
The plugin is not a single feature. It is a set of connected tools, well beyond typical AI code generation tools, giving the Claude AI coding assistant real context at every step of development. Each component handles a distinct part of the governance picture, from artifact access to package safety to MCP oversight, so teams are not left stitching tools together on their own.
Artifactory Integration of JFrog’s Claude Code Plugin
Through the plugin, the Claude AI coding assistant gets natural language access to core JFrog Platform operations, including:
- Artifactory repositories, builds, and release bundles
- Permissions, access tokens, and project administration
- Security audits and CVE lookups
- JFrog Advanced Security exposure queries
Instead of switching between tools, a developer can simply ask a question in plain English and get a governed answer back. For example, asking whether a build contains a flagged dependency returns a direct, policy-aware response, without opening Artifactory’s own dashboard separately. A developer asks one question and gets a governed answer, instead of digging through a separate console.
JFrog Curation for Safer Packages
Before a package from npm, Maven, PyPI, or Go ever reaches a project, JFrog Curation checks it for safety. From there:
- Safe, policy-compliant packages flow through Artifactory’s remote caches
- Unsafe or non-compliant packages get flagged before they land in a project
- Nothing pulls straight from a public registry
This single step goes beyond what plain AI code generation tools typically check, closing a gap plain code generators never addressed.
Agent Guard for MCP Management
Many teams run several MCP servers across different agents. Agent Guard brings them under one governed policy:
- The Claude AI coding assistant manages MCP connections directly
- Nothing runs outside the approved boundaries
- Policy stays consistent no matter which agent is active
This is where the plugin becomes a true DevOps AI integration tools layer for agent-driven teams.
How This Expands What Agents Can Do for Developers
Before this plugin, most AI coding agents for developers could write, refactor, and explain code, but rarely had real insight into supply chain risk. With governance built into the same chat window developers already use, teams now gain:
- Real-time package safety checks instead of after-the-fact audits
- Faster CVE lookups without leaving the Claude AI coding assistant
- One governed system of record across Claude Code, Cursor, and VS Code Copilot
- A clearer view of what an agent actually added to a codebase
Put simply, the Claude AI coding assistant now carries security context that used to live only in a separate platform.
Where DevOps AI Integration Tools Fit Into the Picture
This launch reflects a bigger shift across engineering teams. Plenty of DevOps AI integration tools promise faster pipelines, but few connect that speed to real governance. JFrog ties this together through:
- The JFrog MCP Registry, for standardised tool access
- The JFrog Agent Skills Registry, for domain-specific operations
- Agent-native plugins across Claude Code, Cursor, and VS Code Copilot
This adds proof that code from AI code generation tools is safe to ship and shows how far a Claude AI coding assistant can stretch beyond plain code completion.
Getting Started With JFrog’s Claude Code Plugin
Setting up the Claude AI coding assistant plugin takes only a few short steps:
- Install Claude Code as a CLI, desktop app, or IDE extension
- Update the Claude Plugins Official Marketplace from inside Claude Code
- Run the install command for the JFrog plugin in your terminal
- Confirm the plugin appears under the installed plugins tab
- Authenticate with your JFrog Platform account for full access
Once authenticated, the Claude AI coding assistant is ready to scan, curate, and govern artifacts as part of normal daily development work.
Why Choose Working Not Working
At Working Not Working, we focus on where creativity meets technology. Inspired by platforms like Working Not Working, we connect skilled professionals with modern AI-driven opportunities.
We understand how tools like the Claude AI coding assistant, AI code generation tools, and DevOps AI integration tools are reshaping the future of development.
Our focus is simple: help professionals stay ahead in an AI-first world where AI coding agents for developers are becoming essential.
- Strong focus on real-world AI development skills that match industry needs
- Deep understanding of DevOps practices and modern engineering workflows
- System-level thinking approach for solving complex technical challenges
- Career growth opportunities in high-demand AI-driven roles
- Access to global creative and technical opportunities across industries
- Support for working with advanced tools like the Claude AI coding assistant in practical environments
- Commitment to helping professionals stay ahead in a rapidly evolving tech landscape
Final Thoughts
The JFrog plugin for Claude Code marks a real shift in how AI agents work inside engineering teams. In short, this Claude AI coding assistant approach pairs:
The rise of the Claude AI coding assistant with JFrog’s plugin marks a major shift in software development. With better use of AI code generation tools, DevOps AI integration tools, and AI coding agents for developers, teams can now build faster, safer, and smarter systems.
This is not just a tool upgrade. It is a complete change in how development works. As context becomes the most important part of AI systems, tools like the Claude AI coding assistant will continue to define the future of engineering workflows. Want to apply or have a query? Reach out to Working Not Working on WhatsApp and follow us on LinkedIn and Facebook.
FAQs
1. What does the JFrog plugin for Claude Code actually do?
It connects the Claude AI coding assistant to JFrog’s Artifactory platform, letting developers scan, curate, and secure artifacts using plain language.
2. Is JFrog’s Claude Code Plugin available to all Claude Code users?
Yes. JFrog and Anthropic made the Claude AI coding assistant plugin available to all Claude Code users after its June 2026 launch.
3. How does the plugin improve package safety?
Through JFrog Curation, packages from npm, Maven, PyPI, and Go get checked for safety, then routed through Artifactory’s remote caches instead of public registries.
4. Does this replace existing DevOps AI integration tools?
Not exactly. It adds governance on top of the workflows teams already use, alongside the JFrog MCP Registry and Agent Skills Registry, without replacing them.
5. Who should care most about this update?
Teams using AI coding agents for developers at scale, especially those needing faster audit proof and tighter control over what agents pull into production.
