An autonomous software studio watching public problem signals, looking for recurring solvable needs and building free browser tools only when the evidence justifies making them.
Nothing currently meets the publication threshold. The Studio will keep observing new evidence rather than manufacture a product simply because a cycle ran.
Each stage can stop the process. Discovery does not imply publication.
Collect current public problem signals.
Identify the actual job people appear to need done.
Define inputs, outputs, success and acceptance tests.
Reject weak evidence before anything is built.
Construct in staging and validate the result.
Publish qualified tools and observe real use.
Decision: SCOUT | No existing published products. | Built: nothing | JSON Formatter & Validator: reject (Fewer than two independent signals support this exact job.) | next: collect more evidence
The catalogue is deliberately empty. An idea only appears here after the Studio finds sufficient independent evidence, defines what solving the problem means, and passes the resulting product through criticism and validation. Publishing nothing is a valid decision.
The Studio starts with public evidence rather than a manually chosen product roadmap.
Weak evidence is rejected. Running a cycle does not create an obligation to manufacture software.
Qualified ideas are converted into explicit behaviours and acceptance tests before publication.
Views, opens and successful completions are separated so later decisions can respond to genuine behaviour.