Upstream First logo

Upstream First

Community
mgifford
upstream-first

Load this skill before implementing a fix for a barrier that may originate in a shared component, design system, renderer, or external dependency rather than this project's own code. Decide whether the responsible fix is local, an existing installed capability, or an upstream contribution before writing a workaround. Absolutely always record any temporary local divergence with an owner and removal condition. Under no circumstances open, comment on, or submit anything in an external project without explicit human approval.

Overview

Publishermgifford
Repositoryaccessibility-skills
Skill nameupstream-first
Stars
57
Forks
5
Bundled files
Instructions only
LicenseAGPL-3.0
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 mgifford on GitHub. Read the source before you install it.

Installation

Install the Upstream First 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/mgifford/accessibility-skills/tree/main/skills/upstream-first
  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/mgifford/accessibility-skills.git /tmp/accessibility-skills
mkdir -p .claude/skills
cp -r /tmp/accessibility-skills/skills/upstream-first .claude/skills/upstream-first
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Upstream First 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 Upstream First 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 Upstream First 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.

Upstream First Skill

This skill is a pointer to mgifford/upstream-first, a separate, standalone set of agent skills for deciding where an accessibility (or any other) fix belongs and preparing it for the receiving project. It does not restate that project's decision hierarchy, classification steps, or contribution workflow — those stay owned there so this skill does not drift out of sync with a project that is still a pre-release review draft.

When to Use This

Before implementing a fix for an accessibility barrier, check whether the affected code is actually owned by this project:

  • a shared component or design system used across multiple projects;
  • a renderer, framework, or platform dependency;
  • an npm/PyPI/Packagist/Composer package rather than first-party code;
  • a pattern that would benefit other consumers of the same dependency, not only this project.

If the barrier may originate outside this project's own code, load upstream-first to decide whether the responsible fix is:

  1. an existing installed capability;
  2. an existing public API or extension point;
  3. a suitable maintained open-source package;
  4. a focused contribution to an existing project;
  5. a new reusable open-source package with credible ownership; or
  6. a local implementation, with a written justification for why every higher option is unsuitable.

Do not default to a local patch without going through this decision. A local fix for a shared-component barrier repeats the same maintenance burden for every other consumer of that component, and this project's users are not the only people affected by the barrier.

Preparing a Contribution

If the decision favors an upstream contribution, use maintainer-ready-contribution to prepare an issue, proposal, patch, or pull request the receiving project's maintainers can actually review and accept.

If the issue may be a security vulnerability, do not open a public issue, reproduction, branch, fork, or pull request. Use responsible-security-disclosure instead.

If AI materially assisted a known contribution, run ai-contribution-review before submission, when the receiving project's policy permits that use.

Human Approval Is Required

Never open, comment on, assign, label, fork, branch, or submit anything in an external project without a human explicitly approving that specific action. Deciding that a change belongs upstream is not the same as being authorized to act on that decision in someone else's project.

Tracking Temporary Local Divergence

An urgent local patch, override, fork, pin, or workaround may be necessary while an upstream correction is pending — accessibility barriers should not wait on a release the project does not control. When this happens, do not treat it as resolved. Record it as debt with:

  • a downstream owner;
  • an upstream reference or a documented exception if none exists;
  • a risk-based review date;
  • the supported versions the patch applies to;
  • a removal condition (a supported upstream release adopted and verified); and
  • a rollback plan.

If this project uses the ACCESSIBILITY.md template, use the downstream divergence record. Otherwise use the upstream-debt record shape directly.

A merged or closed upstream tracker item is not resolution. Closure requires the supported release to be adopted downstream and the original user-facing behavior to be retested — see Accessibility Finding Tracking, "Local and upstream trackers" if this project tracks findings with that schema.

Relationship to Other Skills

  • ACCESSIBILITY-general governs when and how all skills in this repository are loaded; this skill is one of the things it points to when a fix's ownership is unclear.
  • bug-reporting records scope, likely source, and (when applicable) local and upstream tracker relationships for a filed finding — this skill decides what to do about that source before a fix is written, not after.
  • Per-topic skills (forms, keyboard, color-contrast, and so on) define how to fix a barrier once ownership is decided. This skill decides where the fix belongs.

References

Frequently asked questions

What does the Upstream First AI skill do?

Load this skill before implementing a fix for a barrier that may originate in a shared component, design system, renderer, or external dependency rather than this project's own code. Decide whether the responsible fix is local, an existing installed capability, or an upstream contribution before writing a workaround. Absolutely always record any temporary local divergence with an owner and removal condition. Under no circumstances open, comment on, or submit anything in an external project without explicit human approval.

Why use Upstream First on TypingMind?

Because you install it once and use it with any model. Upstream First 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 Upstream First in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/mgifford/accessibility-skills/tree/main/skills/upstream-first. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Upstream First?

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 Upstream First?

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

Is the Upstream First AI skill free?

Yes. It is published on GitHub by mgifford under the AGPL-3.0 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 👇