CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

ai-unified-process/aiup-core

AI Unified Process core - stack-agnostic requirements, entity model, and use cases, plus the construction orchestrators for layer plugins

Quality

95%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

stack.mdskills/stack/references/

Stack

For the agent: read before writing code. This file is the AI Unified Process record of the technical layers, independent of how the team documents its architecture. The Layers table says which layer skill builds and tests each technical layer: /use-case-implement runs the Skill column bottom-up, /use-case-test runs the Test skill column from the integration layer up, and /use-case-coverage searches for the markers in the last table.

Last updated: YYYY-MM-DD

Layers

LayerTechnologySkillTest skillLocation
Migration–
Persistence
Client
Service
API
UI
E2E–

Integration layer:

Use case offered throughTested inTest skill
UI
API
Service

A layer the project does not have is marked – in every column (a UI layer in a REST service). The Integration layer table keeps only the rows of the entry layers the project has — a REST service drops the UI row — and its Test skill column names what the Layers table names for that layer (– while the skill is not installed). A use case is tested in the first row its **Layers:** line names, or in the first row when it has no such line. That layer's tests run in process against the real database and exercise every layer below it, so /use-case-test runs its test skill and the ones above it and writes no tests of their own for the layers below unless a run names them (/use-case-test UC-001 persistence).

E2E scope: Use cases and test cases

Use cases and test cases: /use-case-test adds a browser test to every use case offered through the UI, and /test-case-implement automates the TC-* journeys. Test cases only: browser tests are written for the journeys alone, and a use case is tested in its integration layer. A project without an E2E layer drops the line.

External systems

SystemContractLocation

The systems our system calls, by the name the use cases write on their **Secondary Actors:** line: the Client layer builds an adapter for each one from its contract (OpenAPI or AsyncAPI; – while there is none). A secondary actor without a row is a person or a role. Without a Client layer the section is dropped.

Module layout

API contract

Commands

PurposeCommand
Build
Unit tests
Integration tests
End-to-end tests

Markers

Each layer skill keeps its own rows here: on every run it checks that the rows for its layer match its marker reference and corrects them. Do not edit the rows by hand; rerun the layer skill instead.

LayerSkillKindMarker to grep for
All–Code// UC-XXX BR-YYY: directly above the method, query condition, or validator that enforces the rule

skills

README.md

tile.json