implement

User Input

$ARGUMENTS

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

Outline

1. Locate feature directory: Run .specify/scripts/bash/check-prerequisites.sh --json --paths-only from repo root. Parse JSON for FEATURE_DIR, FEATURE_SPEC, and TASKS paths. For single quotes in args, use escape syntax: e.g 'I'''m Groot' (or double-quote if possible).

2. Validate artifacts exist:

   • Check that FEATURE_DIR/spec.md exists. If not: "Run /sdd-change:specify first."
   • Check that FEATURE_DIR/tasks.md exists. If not: "Run /sdd-change:tasks first."

3. Load context:

   • REQUIRED: Read spec.md (the change spec — Problem, Proposed Change, Acceptance Criteria)
   • REQUIRED: Read tasks.md (the flat task list)
   • IF EXISTS: Read .specify/memory/constitution.md to respect project principles
   • IF EXISTS: Read .specify/memory/lessons-learned.md — filter to entries tagged [phase:implement] or [phase:all] and note any relevant to the current change

4. Execute tasks sequentially:

   • Parse all uncompleted tasks (- [ ] items) from tasks.md
   • For each task:
     • Execute the task as described (create, modify, or update files)
     • Follow Rails conventions and the project's architecture (skinny controllers, services for business logic, normalization-only callbacks)
     • Mark the task as [X] in tasks.md immediately after completion
     • Report brief progress: task ID and what was done
   • If a task fails:
     • Provide clear error context (what failed and why)
     • Log the error to .specify/memory/lessons-learned.md if it represents a reusable learning (create the file with its standard header if it does not exist)
     • Halt execution and suggest next steps for manual resolution

5. Final validation:

   • Run bundle exec rspec and report results
   • Run bundle exec rubocop -a and report results
   • If either fails: attempt to fix the issues and re-run (max 2 attempts)
   • If still failing after retries: report the remaining issues for manual resolution