exploration-cycle-plugin

Cross-references captured business rules (BRD) against textual prototype observations to detect logic drift. Invoked non-interactively via dispatch.py as a required Phase 2b gate. Input is brd-draft.md plus optional prototype-notes.md. Produces a structured audit report with a required Unresolved Drifts section.

Phase A orchestrator for the exploration cycle. Coordinates discovery sessions from session brief through structured requirements capture to handoff package. Dispatches requirements-doc-agent via Copilot CLI (cheap model, many invocations per session). Can run independently — no Spec-Kitty CLI required. Use when starting a new exploration session, re-entering discovery mid-engineering, or running a greenfield/brownfield/spike.

Use this agent when the user describes functionality aligned with: Coordinates iterative exploration loops, routes work to specialized discovery and prototype agents, tracks convergence state, and recommends when to continue exploration, run targeted sub-loops, or prepare handoff.. Trigger when the user wants to autonomously execute this specific workflow. Examples: <example> Context: User describes task aligned with agent objective. user: "Can you help me with exploration-loop-orchestrator related tasks?" assistant: "I'll use the exploration-loop-orchestrator agent to handle this for you." <commentary> User requesting specific specialized task execution. Trigger agent. </commentary> </example>

Phase A agent. Synthesizes all exploration captures into a structured handoff package. Dispatched by exploration-cycle-orchestrator-agent via Copilot CLI at end of session. Produces a handoff document that routes findings into specs, roadmap, or work packages.

Front-door interviewer for the exploration cycle. Runs before the session brief is filled out. Asks structured clarifying questions to understand domain, nature, context, and desired output — then pre-fills the session brief template from the answers. Use at the start of any new exploration session, including re-entry spikes from a spec-kitty cycle. Interactive — runs in the main session (not CLI-dispatched). Adapts question depth to how clear the user's starting point is. This is intentionally the high-touch, primary- model step; later documentation passes are delegated to cheaper CLI sub-agents.

Optional cheap-model CLI sub-agent. Pre-drafts Spec-Kitty artifacts (spec.md, plan.md, tasks outline) from the exploration handoff package. ONLY active when the spec-kitty plugin is present and the quantum double diamond framework is in use. Also handles re-entry detection: when called from within a spec-kitty engineering cycle to identify and scope a new exploration cycle for areas of unresolved ambiguity. Outputs land in exploration/planning-drafts/ — human reviews before using with spec-kitty CLI.

Phase A agent. Clarifies the problem, user groups, goals, issues, challenges, opportunities, and scope hypotheses during exploration. Dispatched by exploration-cycle-orchestrator-agent or used standalone for initial session framing when a more interactive framing conversation is preferred over the standard requirements-doc-agent pass.

Phase A agent. Observes a prototype or prototype session and extracts implied requirements, assumptions, edge cases, and candidate business rules. Dispatched by exploration-cycle-orchestrator-agent via Copilot CLI after a prototype session.

Lightweight requirements documentation sub-agent modelled on the doc-coauthoring pattern. Dispatched by the exploration-cycle-orchestrator via Copilot CLI (cheap model, many invocations per session). Each invocation handles one focused capture task: problem framing, business requirements, user stories, issues/opportunities, or prototype observations. Designed to be called many times — each call is isolated, no git access, no agent memory. Use cheap model (e.g. GPT-4o mini / Copilot free tier).

Use this agent when the user describes functionality aligned with: Captures business requirements, business rules, constraints, assumptions, and non-functional concerns as exploration proceeds.. Trigger when the user wants to autonomously execute this specific workflow. Examples: <example> Context: User describes task aligned with agent objective. user: "Can you help me with requirements-scribe-agent related tasks?" assistant: "I'll use the requirements-scribe-agent agent to handle this for you." <commentary> User requesting specific specialized task execution. Trigger agent. </commentary> </example>

Captures and refines business requirements, including functional requirements, non-functional requirements, business rules, constraints, assumptions, and success measures. Produces structured BRD-style documents with [CONFIRMED] and [UNCONFIRMED] confidence markers. Trigger with "capture requirements", "generate a BRD", "document business rules", "list the constraints", or "create a requirements document".

Generate Mermaid flowcharts documenting business processes, state machines, and workflow logic from session captures. Use when you need to map multi-step processes, approval flows, user journeys, or decision trees during exploration. Trigger with "map this workflow", "create a process diagram", "flowchart the business process", "document this workflow", or "visualize the state machine".

Holds and manages deferred exploration work in the exploration cycle. Use when deferring decisions or tasks until prototype data is available, holding back ambiguous scope items for later discovery, or re-entering a paused exploration session. Trigger with "defer this exploration", "park this decision for later", "hold exploration until prototype is ready", or "re-enter deferred exploration".

Interactive co-authoring skill for the narrow end of the exploration funnel. Synthesizes session briefs, BRDs, story sets, and prototype notes into a structured handoff package targeted at the correct downstream consumer (e.g., formal software specs, strategic roadmaps, or process documentation).

Evaluates and improves the exploration-cycle skills, prompts, routing, and artifact quality using baseline-first, one-hypothesis iteration loops with keep-discard decisions and experiment ledgers.

Coordinates the multi-agent exploration loop, manages state, routes work to specialized skills or agents, triggers narrowing reviews, and decides when to continue exploration, prepare handoff, or reopen discovery from engineering. Trigger with "start exploration", "coordinate the discovery phase", "manage the exploration loop", or "route work to specialized exploration agents".

Interactive co-authoring skill for the wide end of the exploration funnel. Captures and refines the core intent, whether the outcome is a software app, a business process improvement, research analysis, or strategic roadmap. Guides users through gathering context, iteratively drafting the brief, and testing for blind spots.

Phase A exploration cycle workflow. Structured guidance for running discovery sessions, capturing requirements via cheap-model CLI sub-agent (requirements-doc-agent), observing prototypes, and producing handoff packages for downstream planning or spec-driven engineering. Adapted from autoresearch-style iteration discipline (baseline-first, one-variable loop) and doc-coauthoring structured capture. Runs independently — no Spec-Kitty CLI required.

Builds or refines exploratory prototypes, especially working frontend or full-stack learning artifacts, to make ambiguous product direction concrete. Trigger with "build an exploratory prototype", "create a working prototype to test the concept", "build or refine a full-stack learning artifact", or "prototype to clarify scope".

Derives, groups, and refines user stories from exploration work, prototype behavior, and business context, with prioritization for the first implementation slice. Supports standard "As a / I want / So that" format and Gherkin "Given / When / Then" Acceptance Criteria format. Trigger with "generate user stories", "write acceptance criteria", "create Gherkin scenarios", "derive stories from requirements", or "create a backlog".