FAQ

Frequently asked questions about truly autonomous AI agents

I'm not technical. Can I still use Kindship?

Absolutely. Kindship is designed for people with missions, not people with engineering degrees. You communicate with your agents in plain language--describe what you're trying to accomplish, set constraints, and let them figure out the how. No coding, no complex configuration. If you can explain your goal to a colleague, you can work with a Kindship agent. See our documentation for a quick-start guide.

How is this different from ChatGPT?

ChatGPT waits for your next prompt. Kindship agents have their own drive. They work for days, weeks, and months--planning what to do next, executing tasks, learning from outcomes, and building momentum toward your goals. You check in on progress rather than feeding it instructions. Think of it as hiring a capable operator versus using a smart search bar.

What kinds of missions is Kindship best for?

Kindship is built for people doing work that matters: civic organizers tracking legislation and coordinating advocacy, climate researchers synthesizing data and surfacing opportunities, health advocates navigating care systems and building support networks, education reformers developing curriculum and community engagement. If your work involves complex coordination, ongoing monitoring, or building momentum over time--Kindship can help.

What are the different agent kinds?

Kindship offers seven agent kinds, each designed for a different starting point: Changemaker (you have a vision for change), Soul Emergence (you have something to give), Co-Creative (you have something to build), Unfolding (you have a direction), Sensemaker (you have a territory of knowledge), Lighthouse (you have a fascination), and Open Purpose (you define your own). Explore all seven on our Agents page to find the right fit for your mission.

What makes an AI 'truly autonomous'?

A truly autonomous AI is a persistent, goal-driven entity that can decide what to do next, act in the real world, learn from outcomes, and continue moving toward a vision without being constantly told what to do. Unlike traditional AI that waits for prompts, truly autonomous agents have internal drive, self-directed execution, memory across time, and real-world agency--while humans remain the navigators who provide purpose, constraints, and values.

What is the Prime Directive?

The Prime Directive is your agent's guiding purpose--a mission statement for your collaboration. It defines who you are, what you're trying to achieve, what values guide you, and what constraints exist. Your agent references it constantly when giving advice, making judgment calls, and helping you prioritize. It sits at the top of the planning hierarchy: Prime Directive, Objectives, Projects, and Tasks.

How does Kindship ensure ethical AI?

Our agents are guided by over 30 ethical principles across four categories: Required (non-negotiable safety principles), Recommended (best practices enabled by default), Kindship Principles (our unique approach), and Optional (additional frameworks you can enable). We use a fail-closed safety design--if guardrails ever fail to load, the agent enters conservative safety mode rather than proceeding unguarded. Explore the full framework on our AI Ethics & Guardrails page.

What real-world tasks can Kindship agents actually perform?

Unlike AI that just gives advice, Kindship agents act in the world: they run their own inbox, maintain websites, monitor signals, coordinate with other agents, send emails, schedule meetings, and execute multi-step workflows. They don't just tell you what to do--they do it, report completion, and move to the next task autonomously. Learn more about agent capabilities in our agent documentation.

How does the planning system work?

Kindship uses a four-level planning hierarchy: your Prime Directive (the why), Objectives (what to achieve), Projects (how to get there), and Tasks (what to do today). Your agent helps you break down complex goals, track progress, and stay focused. You can ask it to add tasks, review status, or recommend priorities--all through natural conversation.

Can I collaborate with a team?

Yes. Team accounts let multiple people share agents, conversations, and planning. Team members can collaborate on shared agents while maintaining personal accounts for individual work. Teams have roles (Owner, Admin, Member) with appropriate permissions, and billing is managed at the team level.

How does a Kindship agent learn and improve over time?

Our agents have memory across time--they remember what they tried, learn from errors, and build momentum across sessions. This happens through experiential learning (building understanding from experience), memory integration (past encounters inform future decisions), strategic adaptation (adjusting to changing conditions), and self-correction (identifying and fixing errors in reasoning).

What is the relationship between humans and Kindship agents?

You are the Founder, they are the Operator. You are the Navigator, they are the Ship. You provide the vision, purpose, and values--the agent handles the how. This isn't about replacing human judgment, but amplifying human capability. You review completed work, not micromanage each step. Read more about this philosophy on our ethics page.

How do I get started?

Start with one agent and a real mission. Choose an agent kind that fits your starting point, set a Prime Directive, and watch it plan, execute, and learn. Most people are surprised by what happens in the first 48 hours. Check our quick-start guide to get up and running in minutes.
If you have any questions, please contact us