ES-105 Evidence¶
Purpose¶
This page defines the evidence expected from ES-105.
Design evidence demonstrates that architecture has been translated into implementation-ready technical decisions.
Required evidence¶
| Artifact | Evidence Purpose |
|---|---|
design_overview.md |
Connects architecture to design work. |
interface_design.md |
Defines component interaction contracts. |
data_design.md |
Defines data structures and evidence fields. |
workflow_design.md |
Defines normal, alternate, and exception flows. |
access_control_design.md |
Defines roles, permissions, and enforcement points. |
ai_interaction_design.md |
Defines AI interaction and human review behavior. |
error_and_exception_design.md |
Defines failure and exception handling. |
technical_decision_records.md |
Records important design decisions. |
design_review.md |
Records design review findings. |
design_readiness_summary.md |
Summarizes readiness for ES-106. |
Evidence quality expectations¶
Design evidence should be traceable to architecture, specific enough for implementation, clear about interfaces and data, explicit about evidence creation, explicit about access-control behavior, explicit about AI review behavior where applicable, reviewable before coding, and useful for testing.
Design quality test¶
Design evidence is sufficient when a reviewer can answer:
- What will be implemented?
- What interfaces exist?
- What data structures are required?
- What workflows are expected?
- What evidence is created?
- What actions require authorization?
- How are failures handled?
- Where does AI appear, if at all?
- What human review is required?
- What decisions were made and why?
AI-use evidence¶
If AI materially assists design work, record how it was used and how output was verified.
Minimum recommended statement:
AI assistance was used to critique design, identify edge cases, or draft decision language. The engineering team reviewed and accepted responsibility for final design decisions.
Review evidence¶
Design should be reviewed before ES-106.
Review should check architecture alignment, implementation clarity, data completeness, interface clarity, workflow coverage, access-control enforcement, evidence preservation, AI-control behavior, failure handling, and testability.
Review findings should be recorded in:
docs/design/design_review.md
Do not treat code as the only design artifact. Important design decisions should be visible before implementation hides them inside code.
Design evidence is the contract between architecture and implementation.