AI Agent Hub
Back to plugins
🖥️

dsh-tool-everything

Client Updated 2026.08.18

Run the following command in DeepSeek Harness:

dsh plugin install zhourenke/dsh-tool-everything

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install zhourenke/dsh-tool-everything in the DeepSeek Harness terminal to install; full source: https://github.com/zhourenke/dsh-tool-everything

About this plugin

Finding a specific file on a Windows machine with an AI model usually means falling back to glob or grep, which can take tens of seconds or more across a million files. dsh-tool-everything taps into the Everything real-time index and plugs it into DSH, so a single everything_search call by the model returns results in milliseconds. Because Everything maintains a live index over every NTFS volume in the background, the search never needs to walk the disk, making it near-instant even across millions of files.

The plugin exposes the full Everything query syntax (wildcards, content:, size:, dm:, ext:, path:, boolean operators) to the model, along with a dozen fine-grained parameters for regex, case sensitivity, whole-word matching, sorting, attribute filtering, and result field selection. The engineering details are equally solid: the GBK-versus-UTF-8 encoding clash on Chinese Windows is resolved by switching the console code page once; shell special characters are escaped with caret to prevent misparsing; es argument order follows its greedy-parsing rule strictly; FILETIME dates, attribute bitmasks, and nested JSON quirks are all normalized in the plugin so the model always receives clean ISO-8601 dates and readable attribute labels.

Built for DSH developers on Windows who want the model to search local files. The only prerequisites are the free Everything install and es.exe on the PATH. If your daily workflow involves locating files among hundreds of thousands of entries by name, size, modification date, content, or path combination, this plugin drops search latency from seconds to milliseconds.

Use Cases

  • Let the model locate a target file among millions of local Windows files in milliseconds
  • Search local files by combining name, size, modification date, and content filters
  • Replace slow glob or grep scans with instant lookups powered by the Everything real-time index

Best For

  • DSH developers on Windows who want the model to search local files
  • Users with large local file sets who frequently need multi-condition file lookup
  • Developers already using Everything and looking to integrate it into an AI workflow