Claude 
Cursor 
Codex 
Hugging Face 
NVIDIA 
Visual Studio Code 
Devin 
Kiro 
Open Code 
NanoClaw 
Docker 
npm 
PyPy 
Wiz 
Finanzdienstleistungen 
Gaming 
Öffentlicher Sektor 
Automobil 
Technologie 
Gesundheitswesen 
Build Trusted Software With Agents That Know Your Supply Chain.
Agents get the organizational context and intelligence from your software supply chain to act on your behalf. Ship software you trust, faster.
What does it mean for coding agents to be supply chain intelligent?
Agents operating with software supply chain intelligence are powered with the policies, context, data, and tool access needed to follow your existing golden path to software delivery. JFrog delivers it through Platform Skills, the JFrog MCP Server, Agent Package Resolution, Agent Guard, and the JFrog Agent Plugin.
Works with popular coding agents & IDEs:
Why Agents Need
Supply Chain Intelligence
Safer Agent Actions
Automatically check agent-triggered installs and platform actions against your policies, so agents use trusted packages and take only approved actions.
Less Context Switching
Provision repositories, run platform tasks, and check security and policy data without leaving your coding session.
Lower Token Spend
Give agents the optimal context to pick the right dependencies and fixes the first time, reducing rework and wasted tokens.
How JFrog Brings Supply Chain Intelligence to Agents
Give agents the right playbook
JFrog Platform Skills give coding agents step-by-step guidance for platform tasks like searching and downloading artifacts, tracing builds, and reviewing security and Curation status.
Ground Decisions in Live Platform Data
The JFrog MCP Server connects compatible coding agents to the JFrog Platform. Agents can pull live package versions, CVE data, Curation status, and build info, so recommendations can be grounded in your actual environment.
Control What Agents Can Use
Agent Guard enforces your organization’s policies natively inside the coding agent, so it can only use the skills, MCP servers, and tools you’ve approved.
Route Agent Installs Through Artifactory
Agent Package Resolution is designed to route agent-triggered installs through Artifactory instead of public registries, so packages and container images can follow the same curated, scanned path your developers already use.
Supply Chain Intelligence, Org-Wide
The JFrog Agent Plugin bundles Platform Skills and MCP tools into one setup, with Agent Package Resolution and Agent Guard available for supported agents, so teams can roll out consistent agent configurations org-wide.
Why Enterprise Teams Choose JFrog
Universal System of Record
Packages, builds, artifacts, security findings, policies, and approved agent assets in one governed place for agents to use.
Multi-Agent Support
Works with Cursor, Claude Code, OpenAI Codex, OpenCode, Kiro, Devin AI, VS Code with Copilot, and JetBrains IDEs.
Agent Governance
Apply your organization’s policies to what agents can access, install, and invoke.
Agent Auditability
Gain visibility into agent activity with a record of the packages agents pull through JFrog and the MCP tools they invoke.
Additional Resources
Frequently Asked Questions
Supply chain intelligence lets a coding agent work within your governed software supply chain rather than around it. With it, an agent can be guided to use your established procedures, access current platform data, create repositories, retrieve dependencies through approved repositories, and operate within tool-access policies.
With JFrog, Skills provide procedural guidance, the MCP Server provides live supply chain context, Agent Package Resolution keeps installs on governed paths, and Agent Guard enforces policy on MCP tool calls. The Agent Plugin bundles these capabilities for supported agents, so you can plan a more consistent rollout across your team. An agent without supply chain intelligence, by contrast, can reach public sources directly, use unapproved tools, and expand your dependency surface without visibility or control.
Use Platform Skills when an agent needs to know how to complete a JFrog task correctly. Skills provide procedural guidance, such as how to search and download artifacts, trace builds, review security and Curation signals, or complete administration tasks.
Use MCP tools when an agent needs current information from JFrog, such as package versions, vulnerability data, build metadata, curation status, or provenance. Skills are the playbook; MCP tools provide the live platform data needed to apply it. Teams can use both through the JFrog Agent Plugin.
The JFrog Agent Plugin brings Platform Skills and MCP tools into one setup, with Agent Guard and Agent Package Resolution available for supported agents. It is designed to reduce fragmented agent setups across your team.
While developers can still configure individual Skills or MCP tools independently, the Agent Plugin provides enterprise teams with the fastest, most integrated path to bring supply chain intelligence to agents org-wide.
Agent Package Resolution is a feature of the JFrog Agent Plugin that governs agent-triggered installs. When a coding agent installs a dependency or pulls a container image the request is designed to route through your Artifactory repositories instead of directly to public registries. JFrog Curation policies determine which packages are permitted, helping agent-triggered installs follow the same governed and scanned paths as installs by human developers.
JFrog offers plugins or integrations for Cursor, Claude Code, OpenAI Codex, OpenCode, Kiro, Devin AI, VS Code with Copilot, and JetBrains IDEs such as IntelliJ. Available capabilities vary by agent.
Securing a coding agent can include governing its supply chain path, in addition to monitoring local developer commands. JFrog helps secure coding agents at install time by routing dependencies through Artifactory with Agent Package Resolution, and at runtime by applying Agent Guard policy checks to MCP tool calls.
