Stack Detection logo

Stack Detection

Community
nahid-sparktales
stack-detection

Establish what a frontend project actually uses — framework, router, package manager, styling, component layer, TypeScript posture — from package.json, lockfiles, config files and the source itself, before loading framework-specific guidance or writing a line of code. Fires at the start of any frontend task in an unfamiliar or half-remembered repo, and whenever you are about to assume a convention. Not for choosing a stack for a new project, and not a substitute for reading the code you are about to change.

Overview

Publishernahid-sparktales
Repositoryagent-dispatcher
Skill namestack-detection
Stars
58
Forks
5
Bundled files
1
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.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by nahid-sparktales on GitHub. Read the source before you install it.

Installation

Install the Stack Detection 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.
    https://github.com/nahid-sparktales/agent-dispatcher/tree/main/skills/frontend/stack-detection
  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/nahid-sparktales/agent-dispatcher.git /tmp/agent-dispatcher
mkdir -p .claude/skills
cp -r /tmp/agent-dispatcher/skills/frontend/stack-detection .claude/skills/stack-detection
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Stack Detection 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 Stack Detection 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 Stack Detection 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.

Stack detection

Most wrong frontend edits are correct code for a stack the project does not use. Detection is cheap; guessing is not.

When this fires

Before the first edit in a frontend repo you have not already mapped in this session, and again whenever you catch yourself about to say "this project probably uses". It does not fire for a repo whose conventions you have already read this session.

Procedure

  1. Find the right package.json first. In a monorepo the root one describes the workspace, not the app. Locate the nearest package.json above the file you are changing, and note whether a root one also exists — dependencies can be hoisted, tooling config usually is.
  2. Read dependencies and scripts together. dependencies and devDependencies say what is installed; scripts say what is actually run. The dev, build and test commands are the ones to use later — do not invent npm run dev if the project spells it differently.
  3. Get real versions from the lockfile, not the range. ^15.0.0 in package.json is a constraint. The lockfile (package-lock.json, pnpm-lock.yaml, yarn.lock, bun.lock*) holds the resolved version, and its filename is also the only reliable statement of which package manager to invoke. Two lockfiles present is a finding, not a choice.
  4. Read the config files that exist, and note the ones that do not. Build/framework config, TypeScript config, lint and format config, test config. Their presence tells you the tool is wired in; their contents tell you path aliases, strictness, and what is excluded.
  5. Confirm the routing and rendering model from the directory layout, not from the framework name. Which route directories exist, where the entry file is, whether routes are file-based or declared in code, and — for frameworks with a server/client split — which files carry the directives that mark the boundary.
  6. Identify the styling layer and its version. A config file's existence does not pin a major version; check the lockfile and how the stylesheet imports the framework. Also check for a CSS modules / styled-in-JS / vanilla CSS layer coexisting with it — mixed styling is common and changes what you should write.
  7. Check for a vendored component layer. components.json or an equivalent marker plus a UI directory means components were copied in and are now project source (see shadcn-ui).
  8. Open two or three real source files in the area you will touch. Config says what is possible; the source says what this team actually does — import style, file naming, test colocation, state library usage.
  9. Write the stack down in one short block, with the file each claim came from, and a separate list of what you could not determine. That list is part of the result.

What a signal does and does not license

  • A framework in dependencies → the framework is installed. It does not tell you which router, rendering mode or directory convention is in use. Read the layout.
  • A config file exists → that tool runs. It does not tell you the major version, and majors move config between files. The lockfile decides.
  • TypeScript installed → files are typed. It does not mean strict is on, that any is discouraged, or that generated types are current. Read tsconfig and one real file.
  • A test framework installed → tests can run. It does not mean they pass, cover this area, or are run in CI. Running them is a separate act.
  • A state or data library installed → it is available somewhere. It does not mean the module you are editing uses it. Grep the directory you are in.
  • components.json present → a component CLI was used at some point. It does not mean the components are unmodified, current, or the only UI layer.
  • A lockfile entry → what would install. node_modules is what actually runs; if behaviour contradicts the lockfile, the installed tree is the tiebreaker.

Checklist

  • The package.json read is the one governing the file being changed
  • Package manager identified from the lockfile, not from habit
  • Framework and styling versions taken from the lockfile
  • Routing/rendering model confirmed from the directory layout
  • Dev, build and test commands quoted from scripts
  • At least two real source files read for convention
  • Unknowns written down rather than filled in by assumption

Failure handling

  • No lockfile — versions are unknown. Say so; do not resolve the range in your head. Behaviour that depends on a major version has to be checked against the installed tree or the docs.
  • Conflicting signals (two lockfiles, a config for a tool that is not installed, a framework version that contradicts the layout) — report the conflict. Picking one silently makes you the source of the next bug.
  • Vendored or patched dependencies (patches/, resolutions/overrides, a checked-in fork) — upstream documentation may not describe what runs here. Read the patch.
  • Detection disagrees with the user — say what you found and where, then follow the user. Do not quietly rewrite their description of their own project.

Evidence to report

A short stack block — package manager, framework and version, router/rendering model, styling layer and version, component layer, TypeScript posture, test setup, dev/build/test commands — with the file each line came from, followed by an explicit "not determined" list. A stack summary with no file references is a guess with formatting.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Stack Detection AI skill do?

Establish what a frontend project actually uses — framework, router, package manager, styling, component layer, TypeScript posture — from package.json, lockfiles, config files and the source itself, before loading framework-specific guidance or writing a line of code. Fires at the start of any frontend task in an unfamiliar or half-remembered repo, and whenever you are about to assume a convention. Not for choosing a stack for a new project, and not a substitute for reading the code you are about to change.

Why use Stack Detection on TypingMind?

Because you install it once and use it with any model. Stack Detection 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 Stack Detection in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/nahid-sparktales/agent-dispatcher/tree/main/skills/frontend/stack-detection. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Stack Detection?

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 Stack Detection?

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

Is the Stack Detection AI skill free?

Yes. It is published on GitHub by nahid-sparktales 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 👇