semble
Fast and Accurate Code Search for Agents. Uses 99% fewer tokens than grep+read
Semble is an AI sub-agent and MCP server that provides fast, accurate code search capabilities for AI coding agents like Claude Code, Cursor, Codex, and OpenCode. It enables agents to retrieve precise code snippets from any codebase using natural language queries while consuming approximately 99% fewer tokens than traditional grep-and-read approaches. The tool runs entirely on CPU with no API keys or external services required, indexing repositories in under a second and answering queries in milliseconds. Developers can integrate Semble as an MCP server, CLI tool via AGENTS.md instructions, or as a dedicated sub-agent.
Key Features
Use Cases
- 01Enable AI coding agents to quickly find authentication flows, error handling, or configuration logic in unfamiliar codebases
- 02Reduce token consumption in AI coding sessions by retrieving only relevant code snippets instead of full files
- 03Search large monorepos or complex projects for API usage patterns, implementation examples, or similar code
- 04Integrate semantic code search into Claude Code, Cursor, Codex, or OpenCode workflows via MCP
- 05Build custom tooling with programmatic code search using the Python library API
- 06Find code similar to a specific file location to understand related implementations or refactoring candidates
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
semble — FAQ
What is Semble?+
Semble is a code search library and AI sub-agent that enables coding agents to retrieve precise code snippets from repositories using natural language queries. It uses 99% fewer tokens than traditional grep-and-read approaches while maintaining high accuracy through semantic and lexical search fusion.
How do I install Semble?+
Install uv, then run 'uv tool install semble' followed by 'semble install'. The interactive installer detects your coding agents (Claude Code, Codex, OpenCode, etc.) and configures MCP server integration, AGENTS.md instructions, or sub-agent files automatically.
Which AI clients and coding agents does Semble work with?+
Semble works with Claude Code, Cursor, Codex, OpenCode, VS Code, and any other MCP-compatible agent. It can be integrated as an MCP server, via AGENTS.md/CLAUDE.md instructions, or as a dedicated sub-agent.
Does Semble require API keys or external services?+
No, Semble runs entirely on CPU with no API keys, GPU, or external services required. The embedding model is downloaded once from Hugging Face and cached locally for offline use.
Is Semble free to use?+
Yes, Semble is open source and released under the MIT license, making it free for both personal and commercial use.
How does Semble achieve 99% token savings?+
Instead of grepping files and reading them in full, Semble returns only the specific code chunks relevant to the query. It uses semantic embeddings and BM25 lexical matching with code-aware ranking signals to identify the exact snippets agents need, dramatically reducing token consumption while maintaining high recall.
How do I install semble?+
Open the source repository on GitHub and follow its README. semble is a agent — MCP Agents Market links you directly to the official repo.
Is semble free?+
semble is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.