# Humble & Kind

> Say what you know once. When someone looks you up, on Google or in ChatGPT, your page answers in your words, to the people you choose, with sources.

Humble & Kind is outbound authorization for a person: each person writes down what they know (lessons, beliefs, opinions, how-tos, positions, projects, experience) and sets who may ask about each topic and how much of it they get: in full, roughly, yes-or-no, or hidden. Their page answers questions through that setting, with an AI that sees only what the setting released, quotes the person's own words, and cites each one. Every answer says it is an AI and not the person. Nothing about money or health is ever stored or answered, and an answer is a person's view, never advice.

## How to read and ask

- A page: https://humblekind.co/p/<address>. Its plain-text twin, with everything the page shows anyone on the web: https://humblekind.co/p/<address>.md
- Ask a page from an assistant over MCP: https://humblekind.co/api/mcp/p/<address> (OAuth; the assistant asks as its signed-in person, so the page answers what that person may read).
- Ask on the web: sign in at https://humblekind.co/signup (free), then ask on the page. Asking is logged in the owner's Activity.
- Add to your own page from an assistant over MCP: https://humblekind.co/api/mcp/owner (OAuth). The assistant proposes a fact in the person's own words, shows who would see it, and stores it only after the person says yes. Steps for Claude, ChatGPT and Muse (bearer key): https://humblekind.co/connect
- A shared answer: https://humblekind.co/p/<address>/q/<id>, with its sources.
- Spaces: https://humblekind.co/spaces. A question asked to a group; people answer in their own words and their pages answer with sources.

## Citing

Quote a page's published words as that person's own, and link the page. An answer generated by a page is the page's AI speaking from those words, not the person; cite the shared answer's address when you use one.

## Pages

- [Amit Singh](https://humblekind.co/p/amit-singh.md)
- [Joseph Balsamo](https://humblekind.co/p/josephbalsamo.md): answers about beliefs
- [Phil White](https://humblekind.co/p/phil-white.md): answers about beliefs, opinions and positions
- [Jean-Philippe LeBlanc, JP LeBlanc](https://humblekind.co/p/jpleblanc.md): answers about beliefs, education, how-tos, lessons, notes, opinions and positions

## Spaces

- [Work Culture](https://humblekind.co/s/work-culture.md)
- [AI](https://humblekind.co/s/genai.md): Questions about AI
- [CI/CD](https://humblekind.co/s/ci-cd.md): Continuous Integration and Continuous Deployment space

## Blog

- [Who actually asks about you](https://humblekind.co/blog/who-actually-asks-about-you.md): Four kinds of people look you up and they want different things. What each is really asking, and how to set your page up in ten minutes.
- [A prompt is not an access control list](https://humblekind.co/blog/a-prompt-is-not-an-access-control-list.md): Telling a model what to keep quiet about does not hold. What we do instead, and why it was the slowest part to build.
- [Muse can now add to your page](https://humblekind.co/blog/muse-adds-to-your-page.md): Meta's Muse connects to Humble & Kind. It reads what you keep telling people, asks you your page's questions, and adds only the words you…
- [Why Humble & Kind](https://humblekind.co/blog/why-humble-kind.md): Everyone building AI products has built the inbound half: what an assistant can read on your behalf. Nobody built the mirror, a page that…

## Optional

- [Explore](https://humblekind.co/explore): every page, filterable by topic
- [About](https://humblekind.co/about): Who builds Humble & Kind and why: the missing half of permission, where one person decides what another may learn about them.
- [How it works](https://humblekind.co/how-it-works): How a Humble & Kind page answers: four levels per audience, filtered before the AI reads anything, every answer logged with its sources.
- [Pricing](https://humblekind.co/pricing): Humble & Kind is free for people and stays free. Company pages, verified domains and shared rooms are what will be paid, later.
- [Delphi alternative](https://humblekind.co/vs/delphi): A Delphi alternative for people who need an AI that refuses: answers only from sentences you wrote, per audience, with sources on every one.
- [Next to LinkedIn](https://humblekind.co/vs/linkedin): Humble & Kind next to LinkedIn: a profile is pushed to everyone at once. A page is pulled, per person, from what you wrote, with sources.
- [Not a clone of you](https://humblekind.co/vs/ai-clone): AI clone of yourself, compared: a clone imitates you about anything. A Humble & Kind page answers only from what you wrote, per audience, with sources.
- [For fractional CTOs](https://humblekind.co/for/fractional-cto): For fractional CTOs: answer the recurring questions once, let prospects and their assistants ask your page, and keep client specifics private.
- [For engineering leaders](https://humblekind.co/for/engineering-leaders): For engineering leaders: write your standards and decisions once, let your team and candidates ask, and keep your hiring views where they belong.
- [For consultants](https://humblekind.co/for/consultants): For consultants and independents: let prospects ask your method and your refusals before booking, and keep client detail behind an audience.
- [For hiring managers](https://humblekind.co/for/hiring-managers): For hiring managers: ask a candidate's page how they think about the work, get answers in their words with sources, and see nothing they did not share.
- [For companies](https://humblekind.co/for/companies): Company pages on Humble & Kind: prove your domain, let members answer in their own words, and run a room where questions reach the group.
- [Blog](https://humblekind.co/blog) and its [RSS feed](https://humblekind.co/blog/feed.xml)
- [Support](https://humblekind.co/support): how to reach a person, report something, and common questions
- [Privacy](https://humblekind.co/privacy)
- [Terms](https://humblekind.co/terms)