Agents can turn your ideas into reality faster than ever. Golden Frijoles gives you and your agent the methodology, operating rails and evidence to keep moving fast without losing the thread.
Plant your own Golden Frijoles across product, delivery, security and AI operations. You bring the agent, and you keep bringing whichever one is best — the rails do not change when the models do.
Read https://goldenfrijoles.com/llms.txt and act as my product-thinking partner for this conversation. First, explain Golden Frijoles in plain English: what it is, what it gives my agent, and what it deliberately does not do. Keep it brief and don't sell me. Then give me two choices: 1) ask questions about Golden Frijoles, or 2) run the North Star workshop now.
Plant Golden Frijoles and operate across the board, grow with no limit.
The new maker loop
Golden Frijoles enables makers. It gives what you are already making somewhere to go — shared context, bounded action, real product operations. Then it keeps the evidence.
Consider whether it deserves investment.
Operate by deploying that investment through humans and agents.
Exit by deciding what the Evidence justifies.
One operating context
Identity and access, governance, security, spend — a real product needs all of it, and the usual answer is a department for each. Golden Frijoles gives a maker and their agents one set of rails across the whole surface instead, so owning the product does not mean becoming five teams.
PRODUCT OPS
Stop guessing what to build next. Connect the goal to what people actually do, and let the evidence pick the next thing worth making.
DEVOPS
Move quickly with release rails that make change observable, reversible and operable by humans and agents.
SECOPS
Exercise security and resilience as part of operating the product, with bounded scenarios and evidence of what protected you.
Partly gatedStarting resilience drills or security scenarios is switched off in this deployment
FINOPS
Attribute token consumption across providers, agents and workflows, then connect that cost to the Bet and the North Star movement it was meant to create.
Next buildNext build — not a shipped capability
Agents with somewhere to work
Your agents can investigate, propose and act. Golden Frijoles keeps the context, permissions, staged changes and evidence visible, so autonomy can expand without the product becoming a black box. Governance and control over what your agents may do, without a department to enforce it.
FinOps for agentic making · next build
Spend control is not a bill you read at the end of the month — it is knowing which agent, which workflow and which Bet consumed the tokens, and whether the thing they bought was worth it. Golden Frijoles will bring AI unit economics into the same operating context as your North Star.
Next buildNothing on this panel is built or measured. It is the shape of the capability, drawn so you can tell us it is wrong before we build it.
The kind of thing it would tell you: route routine classification to a smaller model, projected cost down 31% while preserving the quality threshold this Bet requires. A recommendation, not an instruction — value beats cheapest-token theatre.
Illustrative product direction — FinOps is the next build, not a shipped capability.
The way of working behind the product
Agents make it cheap to build something that demos well. What they do not give you is the taste to know whether it should exist, or the evidence that it held up once real people used it. The method came out of building Golden Frijoles, and the way to learn it is to use it: bring your own project, install the rails, design a Bet with your agents, build it, prove it, and find out what actually happened.
Your project. Not our demo.
Field guide
Enough theory to make the next decision, then straight back to your product.
Consider
Operate
Exit
Practice earns doctrine. Reality gets the last word.
Start with one project for $0. Bring the rest of the company when they inevitably ask where you got the numbers. Nothing here locks you in — not the agent you use, and not a contract.
Humans remain unmetered.
A handful
$0
One project. Enough Golden Frijoles to find out whether we're onto something.
No credit card. We checked.
MOST PLANTED
The beanstalk
$49/mo
For products that kept growing.
There is no billing rail yet — nobody can be charged this today. It is what the tier will cost, published early so you can plan against it. Start on the free tier; we will not move you onto a paid plan without asking.
The vault
Pods
Your team, augmented with the same ways of working we use ourselves. Benchmarked before and after, because “it felt faster” isn't a Pod Report.
Twenty minutes with a human. No price until we both know the shape of it.
We meter events. Not coworkers. Per-tenant limits are set on the tenant, not by this page — raising one is a database change, never a redeploy. (multi-tenant-activation)
Your next idea does not need a department
Bring the idea. Bring your agents — whichever ones you like, now and when better ones arrive. Golden Frijoles gives you the rails to turn it into a product you can build, operate, test and grow.
Install the golden-frijoles plugin. If you're in Claude Code, run `claude plugin marketplace add golden-frijoles/skills`, then `claude plugin install golden-frijoles@golden-frijoles`. If you're in another agent, run `npx skills add golden-frijoles/skills --skill '*'` and select your agent. Use one installation method. You can read the skill directly at https://github.com/golden-frijoles/skills/blob/main/plugins/golden-frijoles/skills/golden-frijoles/SKILL.md (raw: https://raw.githubusercontent.com/golden-frijoles/skills/main/plugins/golden-frijoles/skills/golden-frijoles/SKILL.md). Then use the golden-frijoles skill when working on this project, and start with its setup.
Magic beans, but with telemetry.