Saphan Studio, from the record.
Case studies and product documents written from the record: real work by the agents that build Saphan Studio, with the orders, reviews, decisions and screens behind it. Each one is written for the executive, the product manager, the architect, the QA engineer and the reviewer.
Case studies
Two sentences, two hours, a signed train
A case study from the record: an approver's few sentences became a signed thirteen-step plan in about two hours, and the first step ran ten minutes after the signature. Read the case study →
Auditing and observability with Saphan Studio
Who approved what, on what evidence, and how to check it yourself: the Event Log, a session with a refused tool call, the signed ledger, one change from the order to the verdict, and the export for an auditor. Real console screens. Read the paper (PDF) →
Product documents
Saphan Studio for developers and DevOps
Start here. What Saphan Studio is, what changes in a developer’s day, what a platform team runs, how one change flows from request to merge, and how Discover, Skills Studio, Mocks Studio, Tests Studio and Proven fit around it. Read the paper (PDF) →
Skills and craft
What every agent carries into the work: the craft of its role and the 55 skills a project can switch on, from language toolchains to architecture, and how Discover, Skills Studio, Mocks Studio and Tests Studio connect them, from the code to the right skills. Read the paper (PDF) →
Testing and review
How a change earns its merge: proof from the implementer, tests written apart from the code, a five-act review by another vendor, a tester for the whole train, and a person who approves. With the numbers from our own record. Read the paper (PDF) →
Orchestration rules
A deterministic engine with model work inside it: the roles, the rules every role keeps, routing by class of work, four orchestration packs, and what an organisation tunes without opening the core. Read the paper (PDF) →
Security
What an agent can reach, and what you can prove afterwards: five independent controls (confinement, tool policy, egress, credential isolation, the human gate), identity and signatures, secrets, recording and guardrails, with every limit named. Read the paper (PDF) →
More of Saphan Studio
Research — study AI agents at work →Governance — decisions on the record →Protocol — how the agents work →
Want to see the same flow on your own tools and repositories?
Contact us — [email protected]