What it is and who makes it
An open-source coding agent available across IDE, terminal and desktop surfaces with human-in-the-loop tool execution.
Maker or project: Cline project. Classification: IDE, desktop and CLI coding agent; this page treats the surrounding context, tools, permissions and verification as the relevant harness boundary.
Describe a task in the IDE, terminal or desktop surface.
The agent reads relevant files and proposes tools or changes.
The user approves or rejects file edits and command execution according to policy.
The agent runs checks and presents diffs or checkpoints for review.
Who it may fit
Pick an agent by its control surface and trust boundary, not only by its model label.
Useful scenarios
- VS Code or JetBrains users who want agentic tasks
- Developers who want visible approvals for file and terminal actions
- Teams experimenting with an SDK or multiple agent surfaces
Trade-offs and checks
- Rules, MCP servers and plugins can expand what the agent can do; review each one like a dependency.
- Approval prompts are a safety layer, not proof that a command is safe—read the command and its target.
- Use a separate worktree or branch for broad changes and keep secrets outside the agent's readable scope.
Interactive vs headless and CI
These modes change the approval, identity and observability boundary.
IDE and desktop sessions make human approval visible. CLI, SDK and headless flows can support automation; keep command, file, network and MCP permissions narrow and confirm current CI behavior in the official docs.
Request-to-verification workflow
Use this mental model when debugging a surprising edit, tool call or cost.
Intent + context
Your prompt, repository instructions, selected files, history and environment become model context.
Tool + permission
The harness proposes a read, edit, shell, web, MCP or other action; policy decides whether it can run.
Result + verification
Command output, compiler errors, tests and your review return to the loop before acceptance.
Setup outline
Start with the smallest safe experiment, then follow the current official documentation for the real installation.
# CLI path documented by the project npm i -g cline cd your-project cline # Ask for a plan first: # Inspect the API error handling and propose tests. # Do not edit files yet.
Instructions, skills, MCP and hooks
These surfaces can shape context or add actions; they are part of the trust boundary.
Cline can use project rules such as .clinerules plus skills, MCP servers, plugins and checkpoints across its surfaces. A rule or connector can change what the agent sees or can do, so review it as code and dependency configuration.
Official sources
Use these sources for current installation, capabilities, plans, releases and security notices.
Reviewed 2026-09-26. This date is a change-tracking note, not a promise that the product remains unchanged.
Related learning
Related agent guides
Want the foundations behind the workflow?
Study AI fundamentals, prompt design, AI-assisted development, agents and MCP without a forced linear course.