开源AI工作流 是 AI Skill Hub 本期精选Agent工作流之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
Fast, offline code search and code intelligence for LLM agents,提供快速离线代码搜索和代码智能提示功能,适合LLM代理使用。
开源AI工作流 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
Fast, offline code search and code intelligence for LLM agents,提供快速离线代码搜索和代码智能提示功能,适合LLM代理使用。
开源AI工作流 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
# 方式一:cargo install(推荐) cargo install greplm # 方式二:从源码编译 git clone https://github.com/KhaledSMQ/greplm cd greplm cargo build --release # 二进制在 ./target/release/greplm
# 查看帮助 greplm --help # 基本运行 greplm [options] <input> # 详细使用说明请查阅文档 # https://github.com/KhaledSMQ/greplm
# greplm 配置说明 # 查看配置选项 greplm --config-example > config.yml # 常见配置项 # output_dir: ./output # log_level: info # workers: 4 # 环境变量(覆盖配置文件) export GREPLM_CONFIG="/path/to/config.yml"
<img src="docs/logo.svg" alt="greplm" width="320" />
Your coding agent is reading whole files to find one function. Stop it.
greplm is a hot local index for LLM agents. It answers search, call-graph, go-to-definition, AST, and git queries in milliseconds and hands back the exact lines that matter — never file bodies. Same answers as grep-and-read, with up to ~99% fewer tokens, fully offline.
Install · Quick start · Use it from your agent · Why greplm? · Docs
</div>
curl -fsSL https://raw.githubusercontent.com/KhaledSMQ/greplm/main/install.sh | sh
Or with Rust: cargo install --locked --git https://github.com/KhaledSMQ/greplm greplm-cli greplm-mcp
Prebuilt binaries are on GitHub Releases. See Getting started for Windows, agent file setup, and build-from-source.
Build the index once, then query it as many times as you like — no re-scanning the tree:
greplm index # build the index (incremental on re-runs)
greplm search "SegmentWriter" # search file contents
greplm symbols Searcher # look up definitions
greplm refs SegmentWriter # find references
Every result is a compact, jump-ready location instead of a wall of file text:
$ greplm search "SegmentWriter" --path segment.rs --limit 4
crates/greplm-core/src/segment.rs:107:12: pub struct SegmentWriter {
crates/greplm-core/src/segment.rs:114:6: impl SegmentWriter {
$ greplm symbols Searcher --limit 4
function searcher crates/greplm-core/src/lib.rs:204-206
struct Searcher crates/greplm-core/src/search.rs:351-355
Re-run greplm index after changes (it's incremental), or use greplm watch to keep it fresh.

</div>
ripgrep is the right tool for a human scanning a terminal. greplm is for the agent loop — a queryable index it can hammer thousands of times without ever re-scanning the tree.
| greplm | ripgrep | ctags / LSP | |
|---|---|---|---|
| Content search (literal/regex/word) | ✅ | ✅ | ❌ |
| Symbol definitions (14 languages) | ✅ | ❌ | ✅ |
| Call graph: callers / callees / impact | ✅ | ❌ | ⚠️ LSP only |
| Typed go-to-definition | ✅ | ❌ | ✅ LSP |
| Structural / AST search | ✅ | ❌ | ❌ |
| Git time-travel (blame/history/changed) | ✅ | ❌ | ❌ |
| Task context packs (budgeted) | ✅ | ❌ | ❌ |
| Persistent incremental index | ✅ | ❌ scans each run | ⚠️ regenerate |
| Warm daemon (sub-ms queries) | ✅ | ❌ | ❌ |
| Token-compact output for agents | ✅ | ❌ | ❌ |
| MCP server + ready-made agent files | ✅ | ❌ | ❌ |
| Fully offline / no network | ✅ | ✅ | ✅ |
Full breakdown: Features & comparison.
该项目提供了一个开源的AI工作流,支持快速离线代码搜索和代码智能提示功能,适合开发者使用,评分7.5分
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
经综合评估,开源AI工作流 在Agent工作流赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | greplm |
| 原始描述 | 开源AI工作流:Fast, offline code search and code intelligence for LLM agents.。⭐8 · Rust |
| Topics | workflowrust |
| GitHub | https://github.com/KhaledSMQ/greplm |
| License | MIT |
| 语言 | Rust |
收录时间:2026-06-08 · 更新时间:2026-06-08 · License:MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端