What is Enso Method?
Enso Method is an open-source set of agent skills for using Claude Code and Codex to build production software from stated requirements through implementation and testing. It is for people and teams who want AI coding agents to work from documented product intent rather than a stream of chat instructions. The workflow creates requirements and technical design documents, breaks work into phases, and checks the finished software against acceptance criteria.
Enso Method Features:
Requirements Intake: The skills turn source material such as meeting transcripts or earlier project documents into a record of stakeholder decisions.
PRD Creation: They produce business-language PRDs with testable acceptance criteria and an explicit out-of-scope section.
Document Refinement: Autonomous review rounds find and fix issues in PRDs and technical designs, researching gaps and recording assumptions.
Technical Design: The workflow documents architecture, interfaces, data models, and integration points against the code and project standards.
Phased Implementation: Work is divided into independently shippable phases sized for one AI session, with each phase handed to the agent with its relevant context.
Acceptance Testing: The skills map acceptance criteria to tests that exercise the expected behavior, using real resources where available, and audit the PRD when work is complete.
Enso Method Benefits:
Fewer Rework Loops: Resolving open questions and gaps before implementation can reduce corrections caused by an agent making choices the team did not intend.
Less Active Steering: Once intent and phase boundaries are documented, people can spend less time directing each coding step.
Clearer Decisions: Written requirements and assumptions give stakeholders a shared record of what the software should do and which choices still need confirmation.
Session Continuity: Documents passed between skills help a new agent session pick up the work without relying on chat history.
Evidence of Completion: Testing against each acceptance criterion gives teams a concrete basis for judging whether a phase or project is finished.
Use Cases:
New Product Features: Turn a feature idea described in plain language into a PRD, technical design, implementation phases, and tested software.
Meeting Follow-Up: Convert long stakeholder meeting transcripts into a written record of decisions that can guide a software build.
Large Initiatives: Break a broad initiative into smaller PRDs, each focused on work that can ship independently.
Legacy Project Changes: Prepare a technical design for a change by checking its assumptions against the existing code, data, APIs, and standards.
Acceptance Checks: Audit a completed build to verify that each PRD criterion was implemented and has a test that demonstrates its behavior.