One Integration, Every Agent in the IDE
Because MCP servers are configured once in AI Assistant, JFrog context reaches whichever agent a developer chooses — Junie, Claude Agent, Codex, or GitHub Copilot — rather than requiring a separate plugin per agent. Governance follows the developer instead of the tool.
Governed Platform Access Through the JFrog MCP Server
Registering the JFrog remote MCP endpoint gives agents artifact search, CVE and exposure queries, curation and compliance checks, and build traceability through natural language, authenticated against the JFrog Platform rather than through credentials pasted into a config file.
Package Safety Checks and Curation-Aware Resolution
Before a dependency is added, the package safety skill checks it against your curation policy and resolves it through Artifactory rather than a public registry, so an agent editing a build file cannot quietly introduce a component your policy would block.
Project Steering the Team Can Version
JetBrains agents read persistent instructions from an AGENTS.md file in the project root, so JFrog conventions — which repositories to resolve from, what to check before adding a dependency — live under version control and apply to every task rather than being retyped per prompt.
Governed Agent and MCP Discovery
JetBrains notes that each agent from its registry carries its own license, service, credentials, and privacy terms, and advises checking these before granting repository access. Routing discovery through the JFrog AI Catalog and Agent Guard restricts developers to agents and servers the platform team has approved.