Skip to content

Latest commit

 

History

History
56 lines (43 loc) · 3.63 KB

README.StrongAI.Component.md

File metadata and controls

56 lines (43 loc) · 3.63 KB

The WorkBench requirements linter directory provides a cohesive API for evaluating and improving requirements and user stories. entry.ts serves as the single integration point, re-exporting API contracts and main analysis functions. Evaluate.ts orchestrates prompt selection, guideline loading, model interaction, improvement, and splitting. QuickCheck.ts offers lightweight yes/no classification via compact prompts. PromptIds.ts centralizes stable prompt identifiers. globals.d.ts enables Markdown imports at compile time. External collaborators include a prompt repository and chat driver, OpenAI’s chat API, Node’s filesystem and path for guideline files, and shared API type definitions.

C4Component
title WorkBench Requirements Linter - Component Diagram

Container_Boundary(app, "Workbench Linter Library", "TypeScript", "Linting and improvement toolkit") {
  Component(entry, "entry.ts", "Module", "Unified public API re-export")
  Component(evaluate, "Evaluate.ts", "Module", "Evaluate and improve requirements/user stories; split requirements")
  Component(quick, "QuickCheck.ts", "Module", "Rapid checks for requirement/user story shape")
  Component(promptIds, "PromptIds.ts", "Module", "Stable prompt identifiers")
  Component(globals, "globals.d.ts", "Types", "Ambient .md module declarations")
}

System_Ext(apiTypes, "RequirementsLinterApiTypes", "Types", "Shared request/response interfaces")
System_Ext(promptRepo, "Prompt Repository & Chat Driver", "Library", "Prompt store and model invocation")
System_Ext(openai, "OpenAI API", "Service", "External chat model")
System_Ext(fsnode, "Node.js FS & Path", "Runtime", "Reads guideline markdown files")
System_Ext(mdFiles, "Guideline Markdown Files", "Docs", "Requirements and user story guidelines")
System_Ext(promptsJson, "Prompts.json", "Data", "In-memory prompt templates")

Rel(entry, evaluate, "re-exports evaluateRequirement, evaluateUserStory")
Rel(entry, quick, "re-exports quickCheckLooksLikeRequirement/UserStory")
Rel(entry, apiTypes, "re-exports API types")

Rel(evaluate, promptIds, "uses prompt identifiers")
Rel(quick, promptIds, "uses prompt identifiers")

Rel(evaluate, promptRepo, "loads prompts, runs chat driver")
Rel(quick, promptRepo, "loads prompts, runs chat driver")

Rel(promptRepo, openai, "invokes chat completions")

Rel(evaluate, fsnode, "reads guideline files")
Rel(evaluate, mdFiles, "loads guideline content")

Rel(quick, promptsJson, "reads prompt templates via repository")
Rel(evaluate, promptsJson, "reads prompt templates via repository")

Rel(evaluate, apiTypes, "consumes and produces API shapes")
Rel(quick, apiTypes, "consumes and produces API shapes")

Rel_D(evaluate, globals, "type-only .md import enablement")
Loading

Key components

  • entry.ts: Consolidates public surface area, re-exporting evaluation, quick checks, and canonical API types for simple consumption.
  • Evaluate.ts: Core orchestration; selects prompts, injects guidelines and word counts, calls the chat driver, extracts code-fenced results, and splits requirements into atomic statements.
  • QuickCheck.ts: Fast heuristic classification using lightweight models; normalizes responses and provides friendly fallbacks.
  • PromptIds.ts: Single source of truth for prompt identifiers, ensuring consistency and versioning.
  • RequirementsLinterApiTypes: Establishes stable wire contracts for requests and responses.
  • Prompt Repository & Chat Driver: Loads Prompts.json and executes prompts against the OpenAI API.
  • Node.js FS & Path with Guideline Markdown Files: Supplies guideline content consumed by Evaluate.ts.
  • globals.d.ts: Compile-time enabler for .md imports, no runtime impact.