Cline
A coding agent for reviewable development work in editors, terminals, and Node.js integrations.
- Source repo
- cline/cline
- Stars
- ★ 71k
- Last updated
- today
- License
- Apache-2.0
- Primary language
- TypeScript
- FA score
- 54/100 · Major gaps
At a glance
- How it runs
- Works with
- Universal · cross-platformOpenAI API · Claude API
- Cost
- Free software; you pay for model usage
- Setup effort
- Low · running in minutes
- You'll need
- Typical use
- A developer using VS Code who wants coordinated fixes for missing imports, type mismatches, or syntax errors across several files while reviewing each change.
- Not a fit if
- Teams that need an auditable JetBrains plugin source
- Users who want it working without supplying their own model keys
- Source review
- 54/100 · Major gaps
What does this agent do, and when should you use it?
Cline is an open-source coding agent delivered through a CLI, a VS Code extension, and a Node.js SDK, with an installable JetBrains plugin. It reads project structure, makes coordinated code changes, runs terminal commands, and watches lint and compiler output while working. The CLI supports interactive terminal work as well as headless operation for scripts and CI/CD, while the SDK is intended for custom agents and integrations built on the same engine. Its documented surface also includes MCP servers, plugins, persistent multi-agent teams, scheduled jobs, and connections to Slack, Telegram, Discord, and other messaging platforms. In editors, file edits and terminal commands require approval by default, and changes are exposed as reviewable diffs with checkpoints for modification or rollback.
Cline reads project structure and coordinates changes across files; in VS Code and JetBrains, it presents edits as reviewable diffs and records checkpoints for undo. It runs Bash commands directly, watches output in real time, and can install packages, run builds and tests, deploy applications, manage databases, and continue observing long-running development servers. The CLI accepts prompts or piped input, supports cline --json, and manages MCP servers with cline mcp; cline --team-name starts a multi-agent team coordinated through subtasks, while cline schedule create creates cron-based jobs. The SDK exports items including Agent and createTool for registering custom tools and lifecycle hooks, such as a deploy tool.
- A developer using VS Code who wants coordinated fixes for missing imports, type mismatches, or syntax errors across several files while reviewing each change.
- An engineering team that needs to run tests and repair failures from a terminal, or pipe
git diff origin/maininto an automated change-review task. - A CI/CD maintainer scheduling recurring repository work such as open-PR summaries or dependency checks with cron.
- A Node.js developer building an internal agent with
@cline/sdkand custom deployment, audit, or policy-enforcement tools. - A team that wants a coordinator to split a complex implementation, such as authentication with tests, among specialist agents and preserve team state between sessions.
- An organization that needs to interact with agent sessions from Telegram, Slack, Discord, Google Chat, WhatsApp, or Linear while applying access controls.
How do you install or deploy this agent?
Install the CLI:
npm i -g clineFirst invocation:
cline "Run tests and fix any failures"Install the SDK:
npm install @cline/sdkA credential or compatible API endpoint for the chosen model provider is required before model use, but the supplied material does not document the configuration command, environment variables, or npm/runtime version. The VS Code extension and JetBrains plugin each have Marketplace installation entries.
How do you use this agent?
Give the interactive CLI a task, or pipe input to headless CLI usage. For example:
git diff origin/main | cline "Review these changes for issues"For structured output:
cline --json "List all TODO comments"Manage MCP servers with cline mcp, or start a team task with cline --team-name auth-sprint "Plan and implement user authentication with tests". In an editor workflow, use Plan mode to explore and propose a strategy, then Act mode to execute it; edits and commands require approval by default, though auto-approve can be enabled.
What are this agent's strengths and limitations?
- One shared engine is exposed through a CLI, VS Code extension, JetBrains plugin, and Node.js SDK for interactive development, automation, and custom integrations.
- The CLI supports both interactive terminal work and headless CI/CD flows with piped input and JSON output.
- The editor workflow provides reviewable diffs, checkpoint-based undo, and human approval for file edits and terminal commands.
- It documents model paths across Anthropic, OpenAI, Google, OpenRouter, AWS Bedrock, Azure/GCP Vertex, Ollama, LM Studio, and OpenAI-compatible APIs.
- The SDK supports custom tools and lifecycle hooks, while the CLI documents MCP, team coordination, schedules, and messaging-platform connections.
- The supplied material does not provide the provider credential, endpoint, model-selection, or complete SDK
Agentinitialization steps needed for a working integration. - Editor approval safeguards do not define CLI restrictions: the CLI can run headlessly with zero interaction, so adopters need to set their own automation boundaries.
- Cross-file edits, command execution, deployment, and database management touch the local filesystem and shell; MCP and messaging connections can expand access to external systems.
- Although the JetBrains plugin has a Marketplace installation entry, the material explicitly says its source is not currently open, so it cannot be audited in the same way as the repository CLI or SDK.
- Teams, scheduled jobs, and messaging connections require additional commands, access controls, or credentials, while complete operating and failure-handling procedures are not documented here.
How does this agent compare with similar options?
Cline is not tied to one model provider: the material lists Anthropic, OpenAI, Google, OpenRouter, Vercel AI Gateway, AWS Bedrock, Azure/GCP Vertex, Cerebras, Groq, Ollama, LM Studio, and any OpenAI-compatible API. The documented differentiation is therefore the connected model or gateway rather than a change in Cline's CLI, editor, or SDK delivery modes.
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| Cline This agent | 54 · Major gaps | CLIFree + model costs | ★ 71k | today | TypeScript | OpenAI API · Claude API |
| Qwen Code | 54 · Major gaps | CLIFree + model costs | ★ 29k | today | TypeScript | OpenAI API · Claude API |
| Kilo Code — The Open-Source All-in-One Coding Agent | 55 · Major gaps | IDE extensionFreemium | ★ 28k | today | TypeScript | — |
| Octomind | 70 · Some gaps | CLIFreemium | ★ 557 | 1d ago | Rust | OpenAI API · Claude API |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence shows: README mentions human-in-the-loop approval and auto-approve options, but does not detail least-privilege principles; checkpoints exist for undoing changes, but rollback mechanism details are not provided; SECURITY.md and dependency overrides address known vulnerabilities, but no dependency security audit evidence; data flow transparency and sensitive data handling are not clearly specified. Deductions: insufficient evidence for least privilege, data flow transparency, sensitive data handling, external effects, rollback, and source attribution.
Evidence shows: test files (auth.test.ts, config.test.ts) and CI workflows exist, but no test results provided; dependency overrides and trustedDependencies indicate some management of dependency availability, but no guarantee; failure messages are covered by tests, but no actual failure message examples. Deductions: insufficient evidence for dependency availability and failure messages.
Evidence shows: README describes multiple use cases (CLI, SDK, VS Code, JetBrains, Kanban) and multiple model providers, indicating broad audience; capability boundaries are partially defined via Plan/Act mode and tool system, but not all boundaries listed; trigger precision is partially addressed via CLI commands and scheduling, but not detailed; environment fit is supported across platforms and models, but not all requirements specified. Deductions: incomplete evidence for capability boundaries and trigger precision.
Evidence shows: README provides clear product index and documentation links, good information architecture; installation instructions (npm install) and marketplace links exist; naming is stable (@cline/sdk, cline CLI); example code and FAQ links exist; known limitations not explicitly listed; Apache-2.0 license is clear; CHANGELOG and release process exist; maintenance responsibility is indicated via SECURITY.md and contributing guide. Deductions: known limitations not explicitly stated.
Evidence shows: output usability is supported by JSON output and headless mode for scripting; marginal value is provided by unique features like multi-agent, scheduling, and messaging platform connections; cost-benefit is open-source and free, but no performance or resource consumption data. Deductions: insufficient evidence for cost-benefit.
Evidence shows: README feature claims have some code and test support, but no specific evidence links; cross-source corroboration is provided by multiple products (CLI, SDK, extension) corroborating each other, but no external validation; fact-inference separation is not clearly distinguished in README descriptions. Deductions: insufficient claim traceability and cross-source corroboration.
- Publisher identity is unverified; exercise caution regarding supply chain risks.
- Dependency security has overrides but no full audit report is provided.
- Data flow transparency and sensitive data handling details are insufficient; further review needed.
- Known limitations are not explicitly listed, which may affect expectations in real use.
Recent releases
Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.
-
Release notes excerpt
Added - Cloud tasks: Settings > API Configuration has a Local | Cloud toggle when cloud sessions are on. Cloud picks the Cline model new cloud tasks run on, saved separately from your local models. Clicking the model label in a cloud session's composer opens it there. - Starting or resuming a cloud sandbox shows a progress card with the repo, branch and steps (start sandbox, clone repository, start agent). It collapses once the agent starts. Changed - Cloud sessions hide the Plan/Act toggle and the auto-approve bar instead of showing them disabled. …View on GitHub ↗ -
desktop-v0.0.49 Desktop v0.0.49
Release notes excerpt
- A new Changes button in the session header shows the session's line counts and opens a pane listing the files changed in the last turn or across the whole session, with a diff of each. When a turn finishes, a summary under it lists the files it changed; click it to open the pane on that turn - A project files panel (folder button in the session header) shows a file tree beside the conversation, with a read-only tabbed viewer, git status letters, and a marker on files the session changed. Files the agent edited get a File/Diff toggle. Works on local and SSH environments …View on GitHub ↗ -
cli-v3.0.74 CLI v3.0.74
Release notes excerpt
- Opening a task that's already running in another Cline app now attaches to that run instead of starting a second one that couldn't be stopped. If an app closes mid-task, requests waiting on it now fail with a clear message instead of hanging - cline hub status and cline doctor no longer hang on a Hub that has frozen. Its health check now times out after 3s and the Hub is reported as not running - A turn whose connection drops after the model has streamed only reasoning (for example, "The socket connection was closed unexpectedly") is now retried instead of ending the run …View on GitHub ↗