Peter Yang open-sources an AI skill for cancer patients
Signals InboxAugust 26, 2026Consumer Health
Peter Yang open-sourced an AI skill designed to help cancer patients and caregivers navigate diagnosis, treatment, questions, and self-advocacy. It is a small release, but it captures a bigger behavior change: patients are starting to package lived experience into reusable agent workflows.
Today, I’m open-sourcing /fuck-cancer, an AI skill that helps patients and caregivers navigate cancer diagnosis and treatment and advocate for themselves and their loved ones.
— Peter Yang (@petergyang) August 25, 2026
Here’s what patients and caregivers have told me since I shared my mom’s story:
“You have to be a huge… pic.twitter.com/TwpOM3LE5M
Q1What was actually announced?
Peter Yang released the skill publicly after writing about his mother's repeated breast-cancer treatment. He frames it as a tool for navigating information and preparing for care, not as a replacement for oncologists or a system that makes medical decisions on its own.
Q2How big is the signal?
The software itself is lightweight compared with a regulated medical product. The signal is distribution: reusable AI skills can spread instantly, be forked, and sit on top of general models. A useful patient workflow no longer requires building a full consumer-health app, hiring a large engineering team, or owning a model.
Q3Why does it matter now?
Cancer care is information-dense and emotionally difficult. Patients juggle pathology, treatment options, side effects, appointments, insurance, and second opinions. An agent that organizes questions and records can reduce cognitive load. The pattern could extend to diabetes, fertility, rare disease, and chronic care.
Q4What is the catch?
Medical advice is high stakes. A skill can hallucinate, miss contraindications, or overinterpret a paper, and a patient may trust a confident answer too much. Open-source distribution also means there is no built-in clinical governance. The safest value is preparation and organization, with clinicians still responsible for diagnosis and treatment.
Q5What should we watch next?
Watch whether clinicians contribute to the skill, whether users report concrete workflow benefits, and whether general AI platforms introduce stronger medical guardrails for community-built agents. The bigger question is whether trusted health workflows become a new layer above foundation models.