speckit-plan

User Input

$ARGUMENTS

You MUST consider the user input before proceeding (if not empty).

Spec-Forge Setup (replaces setup-plan.ps1)

1. Read .specify/feature.json for spec_forge_feature_id. If missing, instruct the user to run /speckit-specify first and stop.
2. Call get_feature_snapshot with { featureId } to load the current specification (must exist and be at least in_review) and the active constitution. If the specification is still draft, warn the user and recommend /speckit-clarify or re-running /speckit-specify before planning.

Outline

1. Load context: Use the specification and constitution loaded in Spec-Forge Setup.

2. Execute plan workflow: Draft the plan as a single structured JSON object (this replaces generating research.md, data-model.md, contracts/, quickstart.md as separate files — Spec-Forge stores them as sections of one versioned plan document instead):

   • Fill technicalContext (mark unknowns as "NEEDS_CLARIFICATION: ..." strings)
   • Fill constitutionCheck from the loaded constitution (list each relevant principle and whether the plan complies; ERROR on violations unjustified)
   • Phase 0: Populate research (resolve all NEEDS_CLARIFICATION)
   • Phase 1: Populate dataModel, contracts, quickstart
   • Re-evaluate constitutionCheck post-design (append a second pass entry if anything changed)

Phases

Phase 0: Outline & Research

1. Extract unknowns from technicalContext:

   • For each NEEDS_CLARIFICATION → research task
   • For each dependency → best practices task
   • For each integration → patterns task

2. Generate and dispatch research agents:

   For each unknown in Technical Context:
     Task: "Research {unknown} for {feature context}"
   For each technology choice:
     Task: "Find best practices for {tech} in {domain}"
   

3. Consolidate findings into the plan's research array, one entry per decision:

   { "decision": "...", "rationale": "...", "alternatives": ["..."] }