Claude Code CLI at the tip of your cursor
<p align="center"> <img src="./media/logo.png" alt="CodeSpark" width="240" /> </p> <p align="center"><em>Claude Code CLI at the tip of your cursor</em></p> > Claude Code CLI is great when you can YOLO your way through a project. But in teams and established codebases, decision-making, ownership, and deep understanding matter more. CodeSpark is a different kind of experience — still Claude Code CLI under the hood, but designed to keep you in the driver's seat.  ## Getting started 1. Install and authenticate the [Claude Code CLI](https://code.cursor.com/docs/en/quickstart) 2. Install the CodeSpark extension: [Install in VS Code](https://marketplace.visualstudio.com/items?itemName=codespark.codespark-agent) ## How it works ### Inline agent (`Cmd+I` / `Ctrl+I`) Powered by Claude Code CLI running Haiku, optimized for speed. It works from your cursor, editing the file you're looking at. Most edits are fast, single-turn, file-scoped changes. When the task demands it, the agent reads and writes additional files and goes as wide as it needs — but it always stays within the code, never running commands or reaching outside the project. **Deterministic context retrieval:** - The current file content and cursor position - The closest `CLAUDE.md` in the directory hierarchy (from the file's directory up to the workspace root) - Any files linked from those `CLAUDE.md` files (read into context) - Any directories linked from those `CLAUDE.md` files (expanded as file listings) - The latest research summary from the research agent ### Research agent (`Cmd+Shift+I` / `Ctrl+Shift+I`) Powered by Claude Code CLI running default models, optimized to build your own personal context. It can read files, grep through your codebase, search the web, and fetch documentation — but it cannot edit anything. When the research panel is already open, invoking it again attaches the current file and cursor position as context. The output is inte
Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.