Anthropic Claude Code vs JetBrains AI (2026)

MCP servers connect the agent to tools, databases, APIs, issue trackers and monitoring systems at local, project and user scopes, with OAuth 2.0 sign in for remote servers, claude.ai connectors, and managed settings that allow or deny servers across an organization, alongside built in file, search, shell and web tools.

Sourcecode.claude.com/docs/en/mcp and docs/en/chromeread 2026-10-24

The agent acts through the IDE's own machinery. It edits across files, runs builds, tests and terminal commands, drives the real IDE debugger through a bundled Debugger MCP Toolset and reaches external systems through configured MCP servers, and the IDE itself is exposed as an MCP server to other clients.

Sourcejetbrains.com/help/ai-assistant/junie-agent; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-08-30

Subagents are workers Claude spawns for delegated tasks, and agent teams let a lead agent supervise long running peer sessions that share a task list. Dynamic workflows go further. Claude writes a JavaScript script that runs dozens to hundreds of subagents in the background, up to 1,000 per run and 16 at a time by default, asks before a run starts (every run in Manual mode, the first launch in Auto mode), and the script can be saved and rerun as a command.

Sourcecode.claude.com/docs/en/workflowsread 2026-10-24

Junie autonomously plans and executes complex multi step actions from a single prompt, making large scale edits across files, running tests and terminal commands and using external tools while reporting progress. It works in a plan, execute and verify loop where the plan is an editable document approved before code is written, and the agent iterates when tests fail.

Sourcejetbrains.com/help/ai-assistant/junie-agent and junie.jetbrains.com/docs; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-08-30

Cloud scheduled tasks, in research preview, run a saved prompt against chosen repositories on a schedule, on an HTTP POST to the task's own endpoint, or on GitHub pull request and release events, so they keep working with the laptop closed. GitHub Actions and GitLab CI/CD automate code review and issue triage in CI, GitHub Code Review reviews every pull request automatically, and the agent answers @Claude in Slack and runs from the terminal, desktop, IDEs, the web and headless scripts.

Sourcecode.claude.com/docs/en/routines and docs/en/overviewread 2026-10-24

The same agent engine is invoked from IDE chat, the Junie tool window, a standalone CLI in any terminal, CI/CD pipelines via GitHub Actions and GitLab, and third party ACP clients. Remote and asynchronous runs let a long task started on a workstation be monitored elsewhere.

Sourcejunie.jetbrains.com/docs and jetbrains.com/help/ai-assistant/junie-agentread 2026-08-30

The agent grounds itself in the customer's repository by reading and searching it directly on each run, with project instructions from CLAUDE.md or AGENTS.md and path scoped rules, and MCP connections reach the customer's databases, issue trackers and internal APIs. Context is gathered fresh on each run, and Anthropic does not describe an index, graph or embeddings layer that persists between runs.

Sourcecode.claude.com/docs/en/memory and docs/en/mcpread 2026-10-24

The open file and selection reach Junie as automatic IDE context, with other files attached by @ mention, AGENTS.md guidelines added to every task and MCP servers for outside data. The Junie plugin also has an opt in setting for embeddings based codebase indexing. How that index persists or stays current is not described, so context is assembled per task.

Sourcejetbrains.com/help/ai-assistant/junie-agent and junie.jetbrains.com/docs/junie-plugin-project-settings; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-10-25

Auto memory, on by default in local sessions, holds notes Claude writes itself as it works, such as corrections the user gives and approaches the user confirms. It is scoped per repository and shared across worktrees, loaded into every session through a MEMORY.md index, kept outside the transcript retention sweep until the user or Claude edits or deletes it, and browsable and editable through /memory. CLAUDE.md files are standing instructions the user writes, separate from this memory.

Sourcecode.claude.com/docs/en/memoryread 2026-10-24

Persistent project context is carried in an AGENTS.md instructions file that Junie adds to every task and that can be version controlled, and sessions can be managed and their status checked through slash commands. There is no cross session agent memory or learned state store.

Sourcejetbrains.com/help/ai-assistant/junie-agent; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-08-30

Auto mode is the starting mode for interactive terminal and VS Code sessions from v2.1.283. A separate classifier model reviews each action before it runs and blocks what it judges unsafe, the user's ask and deny rules still apply, and an organization can turn auto mode off in managed settings. Manual mode starts read only and asks before edits, shell commands and network requests, with approve once or allow from then on and fail closed matching for unmatched commands. Plan mode researches and proposes changes without editing, and claude -p and Agent SDK runs start in Manual.

Sourcecode.claude.com/docs/en/permission-modes and docs/en/securityread 2026-10-24

Junie asks permission by default before running bash commands, changing files, performing file operations or using external tools, with approve, reject or always allow per command, and the user can open the diff or the full command in the editor before deciding. Changes can be rolled back per file or all at once, and .aiignore restricts which files the agent may process. Brave mode is an explicit opt in that JetBrains warns can cause data loss.

Sourcejetbrains.com/help/ai-assistant/junie-agent; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-08-30

A SOC 2 Type 2 report and an ISO 27001 certificate sit in the Anthropic Trust Center. Claude Code adds permission modes, sandboxed shell commands with filesystem and network isolation, command injection detection, trust verification for new folders and MCP servers, keychain credential storage, managed settings for organization enforcement and a HackerOne disclosure program. Under commercial terms Anthropic does not train models on Claude Code code or prompts unless the customer opts in.

Sourcecode.claude.com/docs/en/security and docs/en/data-usageread 2026-10-24

User level controls cover .aiignore, which excludes files and directories from agent processing, AGENTS.md, which governs agent behavior under version control, and per command approval gates and .junie/mcp scoping, which limit what the agent may do. JetBrains offers an AI Enterprise tier but does not say what it adds. There is no SOC 2 or ISO attestation specific to the AI product, and no SSO, RBAC or admin governance surface.

Sourcejetbrains.com/help/ai-assistant/junie-agent; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-08-30

OpenTelemetry exports metrics, events and optional traces, with a tool decision event for every permission accept or reject and a tool result event for every completed tool call, and OTEL_LOG_TOOL_DETAILS adds commands, MCP server and tool names and inputs. Telemetry endpoints set in managed settings override developer settings, and all operations in cloud sessions are logged for audit.

Sourcecode.claude.com/docs/en/monitoring-usage and docs/en/securityread 2026-10-24

Junie CLI keeps each session's events log and a readable transcript on disk. It saves the full prompt, response and model usage history for resumable sessions and stores subagent transcripts, so a run can be reconstructed afterwards. These are local files rather than a central audit log. In the IDE, changes stay reviewable as diffs and can be rolled back one by one.

Sourcejunie.jetbrains.com/docs/junie-cli; junie.jetbrains.com/docs/junie-cli.htmlread 2026-10-25

Claude Code runs on the developer's own machine by default, and Remote Control keeps code execution and file access local. Cloud sessions run in isolated Anthropic managed VMs with network access limited by default, or an organization routes them, scheduled tasks included, to self hosted environments on its own infrastructure.

Prompts and outputs travel over TLS 1.2 or later, and at rest the Anthropic API uses AES-256 disk encryption, while Amazon Bedrock, Google Cloud's Agent Platform and Microsoft Foundry apply their own keys, with customer managed keys available on the first two. Commercial data is kept 30 days by default, and Zero Data Retention is turned on per organization for qualified Claude for Enterprise accounts.

Sourcecode.claude.com/docs/en/data-usage and docs/en/securityread 2026-10-24

Junie runs locally in the IDE or terminal, and Junie CLI can use Junie Local, a local inference engine for Apple Silicon, or a local Ollama, LM Studio or LiteLLM server, so code and inference stay on the customer's own machines. The Junie GitHub Action runs entirely on the customer's GitHub runners. JetBrains does not say what organization level hosting options AI Enterprise adds.

Sourcejunie.jetbrains.com/docs/junie-local, custom-llm-ollama and junie-on-github; junie.jetbrains.com/docs/junie-local.htmlread 2026-09-29

Plugins package skills, agents, hooks and MCP servers as one installable unit, and Anthropic's official marketplace is added the first time an interactive terminal session starts, browsable from the /plugin Discover tab or Claude Marketplace on the web, with a context cost estimate for each official plugin and named plugins such as commit-commands. Saved workflows and the bundled /deep-research workflow run as commands, and workflows can ship inside plugins.

Sourcecode.claude.com/docs/en/plugins and docs/en/workflowsread 2026-10-24

Several ready to use coding agents are selectable in AI Assistant chat, namely Junie, Claude Agent, Codex and the Copilot agent, and Junie ships a dedicated code review agent beside its coding agent. Most of these agents do the same job.

Sourcejetbrains.com/help/ai-assistant/agents and junie.jetbrains.com/docs/junie-review-agent; jetbrains.com/help/ai-assistant/agents.htmlread 2026-09-29

Every model Claude Code runs is Anthropic's. The customer picks among Claude models per session or subagent and picks where they are served (the Claude API, Amazon Bedrock, Google Cloud's Agent Platform or Microsoft Foundry), but not the model maker, and Anthropic does not support routing Claude Code to non-Claude models through any gateway.

Sourcecode.claude.com/docs/en/llm-gatewayread 2026-10-24

The model is selected from a list in the chat input with a configurable reasoning level, spanning frontier models from OpenAI, Anthropic, Google and xAI. The CLI adds bring your own key, including OpenRouter, and local models can be served through Ollama, LiteLLM or LMStudio.

Sourcejetbrains.com/help/ai-assistant/junie-agent and junie.jetbrains.com/docs; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-08-30

The Claude Agent SDK gives the same tools, agent loop and context management that power Claude Code as a Python and TypeScript library, with hooks, subagents, MCP, permissions, sessions, skills and plugins, and the CLI can be driven from any language with -p and JSON output.

Sourcecode.claude.com/docs/en/agent-sdk/overviewread 2026-10-24

MCP client support spans the IDE and CLI with project and user scoped mcp.json configs, a registry of pre configured servers and an MCP Installation Assistant. The IDE itself runs as an MCP server exposing IDE tools to external clients including Claude Code, Codex and Copilot CLI, a bundled Debugger MCP Toolset exposes debugger control, and the agent speaks ACP so the same engine drives IDE chat, the tool window and the CLI.

Sourcejunie.jetbrains.com/docs/junie-cli-mcp-configuration and jetbrains.com/help/ai-assistant/mcp; junie.jetbrains.com/docs/junie-cli-mcp-configuration.htmlread 2026-08-30

claude plugin eval runs a suite of realistic prompts against a plugin and scores the results, grading each case with a regex over the reply, a check that a particular tool was called, or a rubric a second model judges. It compares the run against a no plugin baseline and gates CI on the score, so a team can catch regressions when its plugin changes or a new model ships.

Sourcecode.claude.com/docs/en/plugin-evalsread 2026-10-24

Debug mode drives the IDE debugger against the customer's running program to inspect runtime state, and the agent runs tests and terminal commands as it works. That is debugging, with the agent checking its own changes against the customer's existing gates. There is no harness, scored test set or optimization loop for evaluating Junie's behavior, and the benchmark results JetBrains publishes are its own measurements.

Sourcejetbrains.com/help/ai-assistant/junie-agent; jetbrains.com/help/ai-assistant/junie-agent.htmlread 2026-09-29

The Chrome integration lets the agent work in a visible browser window with the user's logins, navigating, clicking, typing, filling forms, uploading files, reading console errors and DOM state and saving screenshots, in Chrome, Edge and other Chromium browsers. Site permissions come from the extension, and the agent pauses for logins and CAPTCHAs. Computer use extends this to native macOS apps.

Sourcecode.claude.com/docs/en/chromeread 2026-10-24

Junie CLI's /demo command launches the app in a Docker virtual machine with Chromium and a desktop. A computer use model drives the interface and records screenshots and video. It is a demo workflow and needs Docker. In the IDE, the agent works through the IDE, terminal and debugger APIs instead.

Sourcejunie.jetbrains.com/docs/junie-cli-demo; junie.jetbrains.com/docs/junie-cli-demo.htmlread 2026-10-25