Blog
Notes from the lab.
technology.feeda().run()
Announcement
Introducing Feeda Technology. Proven at Feeda, now yours.
A home for everything we built to help teams create, test, govern and ship agents and AI apps faster.
Product updates.
Lab OS
Evals with a gate: why every release waits for a Maverick
How we tie automated scoring to a single human sign-off without slowing teams down.
Builder
One session, three models: keeping context when you switch
What we learned moving builds between Claude, Gemini and GPT mid-session.
Feeda Labs
Tokens, OAuth and the cost of an agent nobody watches
The control-plane patterns behind four Feeda properties and one approval rule.
- Lab OS
Introducing task suites for Ferdy’s six modes
- Plugins
Feeda FAQ is now a WordPress plugin
- Builder
Shipping chat.feeda.ai as a PWA in a week
- Feeda Labs
Model budgets per lab, per month
- Lab OS
Writing evals a Maverick can read
- Plugins
Approval Gate: holding agent actions on public pages
- Builder
Living docs: sessions that document themselves
- Feeda Labs
Sharing skills across labs without sharing secrets