Template Authoring logo

Template Authoring

OrganizationPopular
dotnet
template-authoring

Guides creation and validation of custom dotnet new templates from existing projects. Generates a .template.config/template.json that preserves the source project's conventions. USE FOR: creating a reusable dotnet new template from an existing project, bootstrapping .template.config/template.json with correct identity, shortName, parameters, and post-actions, adding parameters or conditional content to a template you are authoring, validating the template.json you are authoring before publishing, packaging templates as NuGet packages for distribution. DO NOT USE FOR: validating an existing template.json as a standalone task (use template-validation), finding or using existing templates (use template-discovery and template-instantiation), MSBuild project file issues unrelated to template authoring, NuGet package publishing (only template packaging structure).

Overview

Publisherdotnet
Repositoryskills
Skill nametemplate-authoring
Stars
5.4K
Forks
416
Bundled files
Instructions only
LicenseMIT
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 dotnet on GitHub. Read the source before you install it.

Installation

Install the Template Authoring 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/dotnet/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/plugins/dotnet-template-engine/skills/template-authoring .claude/skills/template-authoring
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Template Authoring 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 Template Authoring 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 Template Authoring 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.

Template Authoring

This skill helps an agent create and validate custom dotnet new templates. It guides bootstrapping templates from existing projects and validates template.json files for authoring issues before publishing.

When to Use

  • User wants to create a reusable template from an existing .csproj
  • User wants to validate a template.json they are authoring before publishing
  • User is setting up .template.config/template.json from scratch
  • User wants to package a template for NuGet distribution

When Not to Use

  • User wants to find or use existing templates — route to template-discovery or template-instantiation
  • User has MSBuild issues unrelated to template authoring — route to dotnet-msbuild plugin

Inputs

InputRequiredDescription
Source project pathFor creationPath to the .csproj to use as template source
template.json pathFor validationPath to an existing template.json to validate
Template nameFor creationHuman-readable name for the template
Short nameRecommendedShort name for dotnet new <shortname> usage

Workflow

Rules that change the answer

Use these structures exactly; do not invent fields from other template features.

Deliver the requested artifact. When the user says "show", "write", or "give me the content", put the complete JSON/XML in the final response even if you also wrote it to disk. Never edit this skill's SKILL.md or plugin documentation as a substitute for authoring the user's template. Only create or modify template files when the user requested file changes and the target template/project is present.

NeedCorrect structureNever use
Conditional XML in a .csprojXML comments such as <!--#if (database == "SqlServer") --> and <!--#endif --> around the complete elementbare #if lines, which make the XML invalid
Restore generated projectsrestore action 210D431B-A78B-4D2F-B762-4ED3E3EA9025; use primaryOutputs, or args.files containing source-template paths/globsrun-script fields such as executable on the restore action
Restrict to the SDK hosta host constraint whose args is an array containing { "hostname": "dotnetcli" }; its optional version restricts the host/CLI versionusing host version when the requirement is specifically the active SDK version, the invalid host ID dotnet-cli, or unrelated pattern / value fields
Restrict the active SDK versionan sdk-version constraint with a NuGet version/range string in argsa machine-specific exact patch unless the template truly requires it
Preserve CPMkeep generated PackageReference items versionless and package the owning Directory.Packages.props when the template is self-containedadding inline Version attributes
Package templatesa pack project with <PackageType>Template</PackageType> and template content packed below content/describing a layout without showing the requested project file

Step 1: Bootstrap from existing project

Analyze the source .csproj and create a .template.config/template.json:

  1. Copy the source project into a dedicated template-source directory by default, preserving the original project untouched. Modify the original in place only when the user explicitly asks for that layout.
  2. Create .template.config inside the template-source directory.
  3. Generate template.json with identity (reverse-DNS), name, shortName, sourceName (project name for replacement), classifications, and tags
  4. Preserve from source — generic dotnet new templates frequently get these wrong, so verify each is carried over from the original .csproj:
    1. SDK typeMicrosoft.NET.Sdk, Microsoft.NET.Sdk.Web, Microsoft.NET.Sdk.Worker, etc.
    2. Analyzer/package reference metadataPrivateAssets, IncludeAssets, ExcludeAssets
    3. OutputType and other key propertiesTreatWarningsAsErrors, Nullable, LangVersion
    4. CPM participation — no inline Version attributes when a Directory.Packages.props is present
    5. Custom build props/targets and Directory.Build.props conventions
    6. Repo conventions — folder layout, naming, global.json SDK pin

Minimal example:

json
{
  "$schema": "http://json.schemastore.org/template",
  "author": "MyOrg",
  "classifications": ["Library"],
  "identity": "MyOrg.Templates.MyLib",
  "name": "My Library Template",
  "shortName": "mylib",
  "sourceName": "MyLib",
  "tags": { "language": "C#", "type": "project" }
}

Required output — do not stop at a minimal stub. Write the complete .template.config/template.json for the actual source project, then emit a short conventions-preserved confirmation so the user can see nothing was silently dropped. This carry-over is the whole value of the skill; a generic dotnet new template that loses these is why authoring ties with a hand-written stub.

Source .csproj settingCarried over?How
SDK (Microsoft.NET.Sdk.*)template content .csproj uses same SDK
TreatWarningsAsErrors / Nullable / LangVersionpreserved verbatim in template .csproj
PackageReference PrivateAssets / IncludeAssets / ExcludeAssetsmetadata kept on each reference
CPM (Directory.Packages.props present)ManagePackageVersionsCentrally remains enabled and no inline Version attributes are emitted

Mark any row you intentionally omitted as ⚠️ with a reason — never leave it implicit.

Step 2: Validate template.json

Validate the generated template.json using the template-validation skill (it owns the full rule set — required fields, identity format, reserved shortName conflicts, parameter datatypes, post-actions, constraints, and tags).

Quick summary of what gets checked:

  • Required fieldsidentity, name, and shortName must be present.
  • ShortName conflicts — avoid names that collide with dotnet new subcommands. Read the authoritative set from the Commands: section of dotnet new --help for the installed SDK and do not hardcode it (it can change between versions); illustrative examples from current SDKs are install, uninstall, update, list, search, details, create. A conflict happens because dotnet new <name> would be parsed as the subcommand of the same name. Top-level dotnet verbs like build, run, test, and publish do NOT conflict. Run dotnet new list to confirm the name is not already taken.
  • Parameters, post-actions, tags — see template-validation for the complete rules, including the valid datatype list.

Step 3: Refine the template

Based on validation results and user requirements:

  1. Add parameters with appropriate types (string, bool, choice), defaults, and descriptions

  2. Add conditional content using the file type's valid syntax. In XML use template directives inside XML comments, not bare preprocessor lines:

    xml
    <!--#if (database == "SqlServer") -->
    <PackageReference Include="Microsoft.EntityFrameworkCore.SqlServer" />
    <!--#endif -->
    <!--#if (database == "Postgres") -->
    <PackageReference Include="Npgsql.EntityFrameworkCore.PostgreSQL" />
    <!--#endif -->
  3. Configure post-actions for solution add, restore, or custom scripts

  4. Set constraints to restrict which SDKs or workloads the template supports

  5. Add classifications and tags for discoverability

For a restore post-action, prefer primaryOutputs when the project path is known:

json
"primaryOutputs": [{ "path": "MyProject.csproj" }],
"postActions": [{
  "description": "Restore NuGet packages.",
  "manualInstructions": [{ "text": "Run 'dotnet restore'." }],
  "actionId": "210D431B-A78B-4D2F-B762-4ED3E3EA9025",
  "continueOnError": true
}]

If args.files is needed, its paths are matched against the source template before renames, for example "files": ["**/*.csproj"]. Explain that distinction.

Step 4: Test the template locally

For a create-from-existing-project request, this step is required rather than optional: install the authored template, run a dry-run, instantiate it into a temporary output folder, and build the generated project. Report each observed result; inspecting template.json alone does not prove the reusable template works.

bash
dotnet new install ./path/to/template/root
dotnet new mylib --name TestProject --dry-run
dotnet new mylib --name TestProject --output ./test-output
dotnet build ./test-output/TestProject

When packaging is requested, include the complete pack project, not only a directory tree:

xml
<Project Sdk="Microsoft.NET.Sdk">
  <PropertyGroup>
    <PackageId>Contoso.ProjectTemplates</PackageId>
    <PackageType>Template</PackageType>
    <TargetFramework>net8.0</TargetFramework>
    <IncludeBuildOutput>false</IncludeBuildOutput>
    <NoWarn>$(NoWarn);NU5128</NoWarn>
  </PropertyGroup>
  <ItemGroup>
    <Compile Remove="**\*" />
    <Content Include="templates\**\*" Pack="true" PackagePath="content\" />
  </ItemGroup>
</Project>

The pack project's target framework applies only to the content-only packaging project; it does not retarget projects inside templates/. Prefer a broadly available supported framework unless the packaging project itself uses newer build features.

For a self-contained CPM template, the packaged Directory.Packages.props must include <ManagePackageVersionsCentrally>true</ManagePackageVersionsCentrally> and every versionless PackageReference must have a matching PackageVersion. Keep the props file at the intended generated repository root; do not place a duplicate nearer the project where it changes lookup.

Validation

  • template.json passes manual validation with zero errors
  • Template identity and shortName are unique and meaningful
  • All parameters have descriptions and appropriate defaults
  • Template can be installed, dry-run, and instantiated successfully
  • Created projects build cleanly with dotnet build
  • Conditional content produces correct output for all parameter combinations
  • XML template directives are wrapped in XML comments and the generated project parses
  • Host constraints use args[].hostname; restore actions use primaryOutputs or args.files

Common Pitfalls

PitfallSolution
Identity format issuesUse reverse-DNS format (e.g., MyOrg.Templates.WebApi). Avoid spaces or special characters.
ShortName conflicts with CLI commandsAvoid names that match a dotnet new subcommand; read the live set from dotnet new --help and don't hardcode it (illustrative examples: install, uninstall, update, list, search, details, create). Top-level verbs like build/run/test/publish are fine. Run dotnet new list to see if the name is already taken.
Missing parameter descriptionsEvery parameter should have a description and displayName for discoverability.
Not testing all parameter combinationsUse dotnet new <template> --dry-run with different parameter values to verify conditional content works correctly.
Hardcoded versions in templateUse sourceName replacement for project names and consider parameterizing framework versions.
Not setting classificationsAdd appropriate classifications (e.g., ["Web", "API"]) for template discovery.
Reusing fields from a different constraint or post-actionFollow the exact schema: host.args[].hostname, and restore args.files rather than run-script fields.

More Info

Frequently asked questions

What does the Template Authoring AI skill do?

Guides creation and validation of custom dotnet new templates from existing projects. Generates a .template.config/template.json that preserves the source project's conventions. USE FOR: creating a reusable dotnet new template from an existing project, bootstrapping .template.config/template.json with correct identity, shortName, parameters, and post-actions, adding parameters or conditional content to a template you are authoring, validating the template.json you are authoring before publishing, packaging templates as NuGet packages for distribution. DO NOT USE FOR: validating an existing...

Why use Template Authoring on TypingMind?

Because you install it once and use it with any model. Template Authoring 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 Template Authoring in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/dotnet/skills/tree/main/plugins/dotnet-template-engine/skills/template-authoring. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Template Authoring?

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 Template Authoring?

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

Is the Template Authoring AI skill free?

Yes. It is published on GitHub by dotnet under the MIT license. 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 👇