经 AI Skill Hub 精选评估,抓取工具 获评「强烈推荐」。这款Prompt模板在功能完整性、社区活跃度和易用性方面表现出色,AI 评分 8.0 分,适合有一定技术背景的用户使用。
抓取工具 是经过精心设计和反复验证的专业 Prompt 模板集合。这些 Prompt 框架能够有效激活 Claude、ChatGPT 等大型语言模型的深层能力,让 AI 生成更准确、更有价值的输出结果。无需任何安装,直接复制模板内容到 AI 对话框即可使用。
抓取工具 是经过精心设计和反复验证的专业 Prompt 模板集合。这些 Prompt 框架能够有效激活 Claude、ChatGPT 等大型语言模型的深层能力,让 AI 生成更准确、更有价值的输出结果。无需任何安装,直接复制模板内容到 AI 对话框即可使用。
# Prompt 无需安装,直接复制使用 # 支持:Claude / ChatGPT / Gemini / 通义千问 等主流模型 # 使用步骤 # 1. 复制 Prompt 模板内容 # 2. 粘贴到 AI 对话框 # 3. 替换 [占位符] 为实际内容 # 4. 发送后获取结构化输出 # 获取原始文件 git clone https://github.com/johnsellin93/grab
# 粘贴到 Claude/ChatGPT 使用 # 示例 Prompt 结构: 你是一位 [角色],擅长 [领域]。 请根据以下要求完成任务: 任务背景:[描述背景] 具体要求:[详细说明] 输出格式:[期望格式] # 将 [] 内内容替换为实际需求
# grab 配置说明 # 查看配置选项 grab --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 GRAB_CONFIG="/path/to/config.yml"
Required:
zsh ripgrep
Optional: tree tmux wl-copy xclip pbcopy
If tree is not installed, grab --tree falls back to find.
git clone https://github.com/johnsellin93/grab.git
cd grab
chmod +x grab
echo 'export PATH="$HOME/grab:$PATH"' >> ~/.zshrc
source ~/.zshrc
You are not copying results. You are exporting context.
grab turns repository exploration into a deterministic workflow for building AI-ready context from large codebases, one extraction at a time.
It combines ripgrep, sed, and clipboard/tmux workflows into deterministic code-context extraction.
Instead of indexing entire repositories, grab allows developers and AI systems to progressively acquire only the context required for a specific debugging or implementation task.
Unlike repository indexing tools, grab focuses on explicit, user-directed context acquisition.
grab helps developers build explicit repository context for AI systems without relying on repository-wide indexing or fragmented snippets. Context is acquired incrementally as the investigation evolves.
It lets you:
| Command | Purpose |
|---|---|
grab --clear | Reset previous debugging context |
grab --tree | Capture repository structure |
grab ExactPattern | Locate relevant call flows and implementation paths |
grab 500 635 file.cs | Extract an exact implementation range |
grab --functions . Order | Locate Order-related functions |
grab --functions . | index all functions |
Problem statement:
"Users occasionally receive duplicate notifications.
We suspect retry handling might be involved,
but we don't yet know which parts of the codebase to investigate."
grab --functions .
↓
assistant sees function ranges and name context
↓
assistant identifies likely investigation targets
↓
assistant emits batches of grab commands
↓
developer reviews and executes them
↓
repository context expands incrementally to clipboard
The intended workflow is batch-oriented. Rather than extracting one function at a time, the AI generates multiple extraction commands that can be executed together to rapidly expand repository context across related code paths.
Each extraction incrementally expands the active repository context and copies the accumulated result into the active clipboard buffer.
Function indexing provides exact extraction coordinates that assistants can use to request additional surrounding implementation context.
Instead of guessing missing code, the AI proposes deterministic extraction commands that developers can use to progressively acquire explicit repository context.
Example assistant batch emission:
grab --functions .
grab 312 383 NotificationDispatcher.cs ProcessNotificationDelivery
grab 448 486 NotificationDispatcher.cs ShouldRetryNotification
grab 521 564 NotificationDispatcher.cs RecordDeliveryAttempt
grab 612 642 NotificationDispatcher.cs HasRecentSuccessfulDelivery
grab 188 236 RetryPolicy.cs RetryFailedNotification
grab 245 271 RetryPolicy.cs GetRetryBackoffDelay
grab NotificationRetryLimit . # variable / symbol lookup
grab DeliveryDeduplicationWindowMinutes . # variable / symbol lookup
grab "duplicate notification" . # exact text search
The following mappings are not required to use grab, but they significantly improve keyboard-driven AI workflows.
| Mapping | Purpose |
|---|---|
<C-s> | Select the current function or method |
<M-s> | Alternate function-selection mapping |
<M-c> | Copy entire current file to the clipboard |
<M-m> | Indent selected code left by one space |
<M-.> | Indent selected code right by one space |
<Leader>r | Search and replace all occurences of keyword |
<Space>h | Highlight from the current line to a specific line |
p | Paste from clipboard and highlight inserted text |
set clipboard+=unnamedplus
set clipboard+=unnamed
Supported targets:
For fast keyboard-driven AI workflows, browser extensions like Vimium C work well alongside grab.
该工具未明确声明开源协议,商业使用前请联系原作者确认授权范围,避免侵权风险。
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。
AI Skill Hub 点评:抓取工具 的核心功能完整,质量优秀。对于内容创作者和自媒体人来说,这是一个值得纳入个人工具库的选择。建议先在非生产环境试用,再逐步推广。
| 原始名称 | grab |
| 原始描述 | 开源Prompt模板:Terminal workflows for rapid AI-powered debugging and repository analysis。⭐9 · Shell |
| Topics | ai-assisted-developmentcode-navigationcodebase-analysis |
| GitHub | https://github.com/johnsellin93/grab |
| 语言 | Shell |
收录时间:2026-06-11 · 更新时间:2026-06-13 · License:未公布 · 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