Dotnet Cli Distribution logo

Dotnet Cli Distribution

Community
wshaddix
dotnet-cli-distribution

Choosing CLI output format. AOT vs framework-dependent, RID matrix, single-file, dotnet tool.

Overview

Publisherwshaddix
Repositorydotnet-skills
Skill namedotnet-cli-distribution
Stars
79
Forks
13
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by wshaddix on GitHub. Read the source before you install it.

Installation

Install the Dotnet Cli Distribution AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/wshaddix/dotnet-skills.git /tmp/dotnet-skills
mkdir -p .claude/skills
cp -r /tmp/dotnet-skills/skills/dotnet-cli-distribution .claude/skills/dotnet-cli-distribution
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Dotnet Cli Distribution in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Dotnet Cli Distribution on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Dotnet Cli Distribution is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

dotnet-cli-distribution

CLI distribution strategy for .NET tools: choosing between Native AOT single-file publish, framework-dependent deployment, and dotnet tool packaging. Runtime Identifier (RID) matrix planning for cross-platform targets (linux-x64, osx-arm64, win-x64, linux-arm64), single-file publish configuration, and binary size optimization techniques for CLI applications.

Version assumptions: .NET 8.0+ baseline. Native AOT for console apps is fully supported since .NET 8. Single-file publish has been mature since .NET 6.

Out of scope: Native AOT MSBuild configuration (PublishAot, ILLink descriptors, EnableAotAnalyzer, trimming) -- see [skill:dotnet-native-aot]. AOT-first application design patterns (source gen over reflection, DI choices) -- see [skill:dotnet-aot-architecture]. Multi-platform packaging formats (Homebrew, apt/deb, winget, Scoop) -- see [skill:dotnet-cli-packaging]. Release CI/CD pipeline -- see [skill:dotnet-cli-release-pipeline]. Container-based distribution -- see [skill:dotnet-containers]. General CI/CD patterns -- see [skill:dotnet-gha-patterns] and [skill:dotnet-ado-patterns].

Cross-references: [skill:dotnet-native-aot] for AOT compilation pipeline, [skill:dotnet-aot-architecture] for AOT-safe design patterns, [skill:dotnet-cli-architecture] for CLI layered architecture, [skill:dotnet-cli-packaging] for platform-specific package formats, [skill:dotnet-cli-release-pipeline] for automated release workflows, [skill:dotnet-containers] for container-based distribution, [skill:dotnet-tool-management] for consumer-side tool installation and manifest management.


Distribution Strategy Decision Matrix

Choose the distribution model based on target audience and deployment constraints.

StrategyStartup TimeBinary SizeRuntime RequiredBest For
Native AOT single-file~10ms10-30 MBNonePerformance-critical CLI tools, broad distribution
Framework-dependent single-file~100ms1-5 MB.NET runtimeInternal tools where runtime is guaranteed
Self-contained single-file~100ms60-80 MBNoneSimple distribution without AOT complexity
dotnet tool (global/local)~200ms< 1 MB (NuGet).NET SDKDeveloper tools, .NET ecosystem users

When to Choose Each Strategy

Native AOT single-file -- the gold standard for CLI distribution:

  • Zero dependencies on target machine (no .NET runtime needed)
  • Fastest startup (~10ms vs ~100ms+ for JIT)
  • Smallest binary when combined with trimming
  • Trade-off: longer build times, no reflection unless preserved
  • See [skill:dotnet-native-aot] for PublishAot MSBuild configuration

Framework-dependent deployment:

  • Smallest artifact size (only app code, no runtime)
  • Users must have .NET runtime installed
  • Best for internal/enterprise tools where runtime is managed
  • Can still use single-file publish for convenience

Self-contained (non-AOT):

  • Includes .NET runtime in the artifact
  • Larger binary than AOT but simpler build process
  • Full reflection and dynamic code support
  • Good compromise when AOT compat is difficult

dotnet tool packaging:

  • Distributed via NuGet -- simplest publishing workflow
  • Users install with dotnet tool install -g mytool
  • Requires .NET SDK on target (not just runtime)
  • Best for developer-facing tools in the .NET ecosystem
  • See [skill:dotnet-cli-packaging] for NuGet distribution details

Runtime Identifier (RID) Matrix

Standard CLI RID Targets

Target the four primary RIDs for broad coverage:

RIDPlatformNotes
linux-x64Linux x86_64Most Linux servers, CI runners, WSL
linux-arm64Linux ARM64AWS Graviton, Raspberry Pi 4+, Apple Silicon VMs
osx-arm64macOS Apple SiliconM1/M2/M3+ Macs (primary macOS target)
win-x64Windows x86_64Windows 10+, Windows Server

Optional Extended Targets

RIDWhen to Include
osx-x64Legacy Intel Mac support (declining market share)
linux-musl-x64Alpine Linux / Docker scratch images
linux-musl-arm64Alpine on ARM64
win-arm64Windows on ARM (Surface Pro X, Snapdragon laptops)

RID Configuration in .csproj

xml
<!-- Set per publish, not in csproj (avoids accidental RID lock-in) -->
<!-- Use dotnet publish -r <rid> instead -->

<!-- If you must set a default for local development -->
<PropertyGroup Condition="'$(RuntimeIdentifier)' == ''">
  <RuntimeIdentifier>osx-arm64</RuntimeIdentifier>
</PropertyGroup>

Publish per RID from the command line:

bash
# Publish for each target RID
dotnet publish -c Release -r linux-x64
dotnet publish -c Release -r linux-arm64
dotnet publish -c Release -r osx-arm64
dotnet publish -c Release -r win-x64

Single-File Publish

Single-file publish bundles the application and its dependencies into one executable.

Configuration

xml
<PropertyGroup>
  <PublishSingleFile>true</PublishSingleFile>
  <!-- Required for single-file -->
  <SelfContained>true</SelfContained>
  <!-- Embed PDB for stack traces (optional, adds ~2-5 MB) -->
  <DebugType>embedded</DebugType>
  <!-- Include native libraries in the single file -->
  <IncludeNativeLibrariesForSelfExtract>true</IncludeNativeLibrariesForSelfExtract>
</PropertyGroup>

Single-File with Native AOT

When combined with Native AOT, single-file is implicit -- AOT always produces a single native binary:

xml
<PropertyGroup>
  <PublishAot>true</PublishAot>
  <!-- PublishSingleFile is not needed -- AOT output is inherently single-file -->
  <!-- SelfContained is implied by PublishAot -->
</PropertyGroup>

See [skill:dotnet-native-aot] for the full AOT publish configuration including ILLink, type preservation, and analyzer setup.

Publish Command

bash
# Framework-dependent single-file (requires .NET runtime on target)
dotnet publish -c Release -r linux-x64 /p:PublishSingleFile=true --self-contained false

# Self-contained single-file (includes runtime, no AOT)
dotnet publish -c Release -r linux-x64 /p:PublishSingleFile=true --self-contained true

# Native AOT (inherently single-file, smallest and fastest)
dotnet publish -c Release -r linux-x64
# (when PublishAot=true is in csproj)

Size Optimization for CLI Binaries

Trimming (Non-AOT)

Trimming removes unused code from the published output. For self-contained non-AOT builds:

xml
<PropertyGroup>
  <PublishTrimmed>true</PublishTrimmed>
  <TrimMode>link</TrimMode>
  <!-- Suppress known trim warnings for CLI scenarios -->
  <SuppressTrimAnalysisWarnings>false</SuppressTrimAnalysisWarnings>
</PropertyGroup>

AOT Size Optimization

For Native AOT builds, size is controlled by AOT-specific MSBuild properties. See [skill:dotnet-native-aot] for the full configuration. Key CLI-relevant properties include StripSymbols, OptimizationPreference, InvariantGlobalization, and StackTraceSupport.

Size Comparison (Typical CLI Tool)

ConfigurationApproximate Size
Self-contained (no trim)60-80 MB
Self-contained + trimmed15-30 MB
Native AOT (default)15-25 MB
Native AOT + size optimized8-15 MB
Native AOT + invariant globalization + stripped5-10 MB
Framework-dependent1-5 MB

Practical Size Reduction Checklist

  1. Enable invariant globalization if the tool does not need locale-specific formatting (InvariantGlobalization=true)
  2. Strip symbols on Linux/macOS (StripSymbols=true) -- keep separate symbol files for crash analysis
  3. Optimize for size (OptimizationPreference=Size) -- minimal runtime performance impact for I/O-bound CLI tools
  4. Disable reflection where possible -- use source generators for JSON serialization ([skill:dotnet-aot-architecture])
  5. Audit NuGet dependencies -- each dependency adds to the binary; remove unused packages

Framework-Dependent vs Self-Contained Trade-offs

Framework-Dependent

bash
dotnet publish -c Release -r linux-x64 --self-contained false

Advantages:

  • Smallest artifact (1-5 MB)
  • Serviced by runtime updates (security patches applied by runtime, not app rebuild)
  • Faster publish times

Disadvantages:

  • Requires matching .NET runtime on target
  • Runtime version mismatch causes startup failures
  • Users must manage runtime installation

Self-Contained

bash
dotnet publish -c Release -r linux-x64 --self-contained true

Advantages:

  • No runtime dependency on target
  • App controls exact runtime version
  • Side-by-side deployment (multiple apps, different runtimes)

Disadvantages:

  • Larger artifact (60-80 MB without trimming)
  • Must rebuild and redistribute for runtime security patches
  • One artifact per target RID

Publishing Workflow

Local Development

bash
# Quick local publish for testing
dotnet publish -c Release -r osx-arm64

# Verify the binary
./bin/Release/net8.0/osx-arm64/publish/mytool --version

Producing Release Artifacts

bash
#!/bin/bash
# build-all.sh -- Produce artifacts for all target RIDs
set -euo pipefail

VERSION="${1:?Usage: build-all.sh <version>}"
PROJECT="src/MyCli/MyCli.csproj"
OUTPUT_DIR="artifacts"

RIDS=("linux-x64" "linux-arm64" "osx-arm64" "win-x64")
# Note: Native AOT cross-compilation for ARM64 on x64 requires platform toolchain
# See [skill:dotnet-cli-release-pipeline] for CI-based cross-compilation setup

for rid in "${RIDS[@]}"; do
  echo "Publishing for $rid..."
  dotnet publish "$PROJECT" \
    -c Release \
    -r "$rid" \
    -o "$OUTPUT_DIR/$rid" \
    /p:Version="$VERSION"
done

# Create archives
for rid in "${RIDS[@]}"; do
  if [[ "$rid" == win-* ]]; then
    (cd "$OUTPUT_DIR/$rid" && zip -q "../mytool-$VERSION-$rid.zip" *)
  else
    tar -czf "$OUTPUT_DIR/mytool-$VERSION-$rid.tar.gz" -C "$OUTPUT_DIR/$rid" .
  fi
done

echo "Artifacts in $OUTPUT_DIR/"

Checksum Generation

Always produce checksums for release artifacts:

bash
# Generate SHA-256 checksums
cd artifacts
shasum -a 256 *.tar.gz *.zip > checksums-sha256.txt

See [skill:dotnet-cli-release-pipeline] for automating this in GitHub Actions.


Agent Gotchas

  1. Do not set RuntimeIdentifier in the .csproj for multi-platform CLI tools. Hardcoding a RID in the project file prevents building for other platforms. Pass -r <rid> at publish time instead.
  2. Do not use PublishSingleFile with PublishAot. Native AOT output is inherently single-file. Setting both is redundant and may cause confusing build warnings.
  3. Do not skip InvariantGlobalization for size-sensitive CLI tools. Globalization data adds ~25 MB to AOT binaries. Most CLI tools that do not format locale-specific dates/currencies should enable InvariantGlobalization=true.
  4. Do not distribute self-contained non-trimmed binaries. A 60-80 MB CLI tool is unacceptable for end users. Either trim (PublishTrimmed), use AOT, or distribute as framework-dependent.
  5. Do not forget to produce checksums for release artifacts. Users and package managers need SHA-256 checksums to verify download integrity. See [skill:dotnet-cli-release-pipeline] for automated checksum generation.
  6. Do not hardcode secrets in publish scripts. Use environment variable placeholders (${SIGNING_KEY}) with a comment about CI secret storage for any signing or upload credentials.

References

Frequently asked questions

What does the Dotnet Cli Distribution AI skill do?

Choosing CLI output format. AOT vs framework-dependent, RID matrix, single-file, dotnet tool.

Why use Dotnet Cli Distribution on TypingMind?

Because you install it once and use it with any model. Dotnet Cli Distribution is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Dotnet Cli Distribution in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/wshaddix/dotnet-skills/tree/master/skills/dotnet-cli-distribution. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Dotnet Cli Distribution?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Dotnet Cli Distribution?

As many as you like. As long as a model supports skills, you can use Dotnet Cli Distribution with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Dotnet Cli Distribution AI skill free?

It is published on GitHub by wshaddix. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇