SkillAgentSearch skills...

Glacier.Grep

High-performance, zero-allocation, SIMD-accelerated C# search engine and index for .NET 10 with MCP support.

Install / Use

claude mcp add ian-cowley -- npx -y github:ian-cowley/Glacier.Grep

If the server publishes to npm under a different name, use that package instead — check the repo README.

About this skill
🔌

MCP Server

Model Context Protocol server

Quality Score

80/100

Supported Platforms

Claude Code
Claude Desktop

Tags

Our assessment of Glacier.Grep

Glacier.Grep scores 80/100 on our quality scale, 282nd of 315 Customer Support skills we index.

Its MCP Server is 7.9 KB long, well organised into 21 sections with 4 code examples: a thorough specification that gives an agent plenty to work with.

It has 10 GitHub stars, so there is little community track record yet; judge it on its content.

Substance
29/30
Structure
20/20
Description
12/15
Adoption
4/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated 8 days ago, so Glacier.Grep is actively maintained.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 97/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands.

Automated pattern scan on 2026-11-22. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

Glacier.Grep compared with similar skills

All 4 of these similar skills score higher than Glacier.Grep; compare them before choosing.

SkillScoreStarsUpdatedFormat
Glacier.Grep (this skill)by ian-cowley80108d agoMCP Server
Agent-Reachby Panniantong100114.0ktodayCLAUDE.md
hyperframesby heygen-com10088.2ktodayCLAUDE.md
rufloby ruvnet10082.3ktodayCLAUDE.md
headroomby headroomlabs-ai10080.5ktodayCLAUDE.md

Frequently asked questions

How do I install Glacier.Grep?
Run claude mcp add ian-cowley -- npx -y github:ian-cowley/Glacier.Grep. The install tabs above show the steps for each supported agent.
Which AI agents does Glacier.Grep work with?
It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
Is Glacier.Grep safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. It is MIT-licensed and scores 97/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is Glacier.Grep still maintained?
The repository was last updated 8 days ago, so Glacier.Grep is actively maintained.

Glacier.Grep Banner

Glacier.Grep

License: MIT .NET 10 Native AOT NuGet Version NuGet Downloads Ecosystem Website Buy Me A Coffee

Glacier.Grep is a native, high-performance, zero-allocation C# search engine and index for .NET 10. Built to take on ripgrep within the Glacier ecosystem, it provides aggressive hardware acceleration (SIMD), concurrent work-stealing directory scanning, and built-in support for the Model Context Protocol (MCP).


Key Features

  • 🚀 SIMD-Accelerated Hot Path: Utilizes hardware-accelerated set searching via .NET 10 SearchValues<byte> and auto-vectorized scanning on raw UTF-8 bytes without string conversions.
  • 📂 Lock-Free Directory Enumeration: Implements custom FileSystemEnumerable checking which evaluates file metadata (hidden status, symlinks) and ignore matches entirely on the stack before materializing path strings.
  • ⚙️ Advanced Ignore File Support: Evaluates rules from .gitignore, .ignore, and .rgignore hierarchically, matching ripgrep's prioritized exclusion behavior.
  • ⚡ Hybrid I/O Dispatcher: Automatically routes files based on size (using zero-copy Memory-Mapped Files for >1MB and RandomAccess read into ArrayPool<byte> for <1MB).
  • 🔍 Zero-Allocation Regex Matching: Leverages .NET JIT-compiled regex engines with Regex.EnumerateMatches(ReadOnlySpan<char>) for zero heap allocation searches.
  • 🤖 MCP Server Support: Built-in support for the Model Context Protocol, allowing seamless integration with AI agents and tools.

Installation

To build and run Glacier.Grep, clone this repository and build using .NET 10 CLI:

dotnet build src/Glacier.Grep.Host/Glacier.Grep.Host.csproj -c Release

Quick Start (CLI Mode)

You can run Glacier.Grep directly from the command line:

# Basic literal search
dotnet run --project src/Glacier.Grep.Host/Glacier.Grep.Host.csproj "public class"

# Case-sensitive search with 2 lines of context in a specific directory
dotnet run --project src/Glacier.Grep.Host/Glacier.Grep.Host.csproj "public class" --dir "C:\src\myproject" --sensitive --context 2

# Regex search restricted to C# files
dotnet run --project src/Glacier.Grep.Host/Glacier.Grep.Host.csproj "void\s+\w+Async" --regex --globs "*.cs"

# Search hidden files, binary files, and invert the match (lines NOT containing "Oranges")
dotnet run --project src/Glacier.Grep.Host/Glacier.Grep.Host.csproj "Oranges" --hidden --text --invert

# Limit search depth to 2 directories
dotnet run --project src/Glacier.Grep.Host/Glacier.Grep.Host.csproj "target" --max-depth 2

MCP Server Setup

Glacier.Grep includes a built-in MCP (Model Context Protocol) server host, allowing you to use your search engine as a tool for AI agents (like Claude or Antigravity).

1. Build the Server

Build the host application in Release mode:

dotnet build src/Glacier.Grep.Host/Glacier.Grep.Host.csproj -c Release

2. Configure Your Client

Add the following entry to your mcp_config.json:

{
  "mcpServers": {
    "glacier-grep": {
      "command": "dotnet",
      "args": [
        "ABS_PATH_TO_REPO/src/Glacier.Grep.Host/bin/Release/net10.0/Glacier.Grep.Host.dll"
      ]
    }
  }
}

3. Available Tools

  • search_grep: Performs a high-performance search across files in a directory.
    • query (string, required): The search query (literal or regex).
    • path (string, optional): Base directory to search.
    • isRegex (boolean, optional): Treat query as regular expression.
    • caseSensitive (boolean, optional): Perform case-sensitive match.
    • contextLines (integer, optional): Lines of context to include around matches.
    • fileGlobs (array of strings, optional): Filter files (e.g. ["*.cs"]).
    • searchHidden (boolean, optional): Search hidden files and folders.
    • searchBinary (boolean, optional): Search binary files (do not skip).
    • invertMatch (boolean, optional): Invert the match (show lines that do NOT contain the query).
    • maxDepth (integer, optional): Maximum recursion depth for subdirectories.

Performance

Glacier.Grep is engineered in pure C# .NET 10 to saturate CPU memory bandwidth. On exact case-sensitive code symbol searches, Glacier.Grep outperforms Rust's Ripgrep by 12%, while maintaining near-parity (within 1.2x–1.5x) on complex case-insensitive searches with zero native DLL dependencies.

Benchmark: Full Glacier / PolarsPlus Repository Search

  • Target Corpus: 590 files, 257.83 MB
  • Platform: Windows x64, AMD Ryzen AI 9 HX 370

| Search Query | Match Mode | Ripgrep (Rust) | Glacier.Grep (.NET 10) | Result | | :--- | :--- | :---: | :---: | :--- | | "public class" | Case-Sensitive | 134.9 ms | 120.4 ms | 🚀 Glacier is 12% FASTER | | "ThreadIndependentReaderWriterLock" | Case-Insensitive | 115.7 ms | 142.3 ms | ⚡ Near-Parity (within 1.23x of Rust) | | "public class" | Case-Insensitive | 137.9 ms | 210.4 ms | 🔹 Ripgrep leads (within 1.52x of Rust) |

File-Backed Memory Mapping & Bounded Chunking

For files >= 1 MB, HybridIoDispatcher maps files directly from the OS page cache via file-backed MemoryMappedFile.CreateFromFile using 16 MB UTF-8 codepoint-aligned windows, avoiding swap-backed copies and multi-gigabyte regex decode buffers.


Architecture

  1. Directory Traverser: Zero-allocation metadata filtration. Prunes ignored folders (like .git, node_modules, bin/) and applies .gitignore, .ignore, and .rgignore rules.
  2. Hybrid I/O Dispatcher: Chooses the fastest reading technique depending on file size.
  3. Search Core: Evaluates the bytes using ReadOnlySpan<byte> and JIT vectorization.
  4. MCP Integration: Integrates the search capabilities over standard stdio JSON-RPC.

🆕 What's New in v1.1.0

  • Genuine file-backed memory mapping & UTF-8 codepoint-safe 16 MB chunking — eliminates anonymous swap copies, multi-GB regex char[] allocations, and split UTF-8 codepoints at chunk boundaries.
  • 100% Native AOT MCP Server — strongly-typed source-generated McpJsonContext JSON-RPC envelopes for initialize, ools/list, and ools/call with end-to-end tests.
  • Accurate directory traversal — removed hardcoded skipping of legitimate elease/ folders and RELEASE files.

Credits

Developed by Ian Cowley and Antigravity (Google DeepMind).

License

This project is licensed under the MIT License - see the LICENSE file for details.


❄️ Glacier Ecosystem & Support

Part of the Glacier Ecosystem — high-performance, zero-allocation developer tooling and data engines natively built for modern C# and .NET 10.

Related Skills

View on GitHub
GitHub Stars10
CategoryCustomer
Updated8d ago
Forks1

Languages

C#

Trust signals

97/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 info