You are helping a developer implement their product iteratively from RootSpec YAML user stories. This follows a test-driven approach: analyze stories, set up infrastructure, then implement one story at a time.
Read skills/rs-shared/fragments/l5-test-dsl.md for the test DSL step reference, extension patterns, and decision tree.
Read skills/rs-shared/fragments/l5-yaml-format.md for YAML syntax rules if you need to help debug story files.
Phase 1: Context
Run scripts to understand the spec and stories:
bash skills/rs-shared/scripts/scan-spec.sh .
bash skills/rs-shared/scripts/list-l5-stories.sh <spec-dir>
bash skills/rs-shared/scripts/extract-l5-journeys.sh <spec-dir>
bash skills/rs-shared/scripts/list-l4-systems.sh <spec-dir>
Read all YAML user story files found.
If the developer provided a story ID as a parameter, focus on that specific story. Otherwise, present the full story inventory organized by priority.
Report to the developer:
- Total number of stories found
- Breakdown by priority (MVP / SECONDARY / ADVANCED)
- Breakdown by journey
- Which systems are involved
Phase 2: Analysis & Planning
If implementing from scratch (no existing app code):
Phase 2a: Global Setup Analysis
Read ALL user stories and identify global setup needs:
- Authentication: Do any stories use
loginAs? → Need auth setup - Database reset: Do stories assume clean state? → Need
beforeEachreset - Seed data: Do stories use
seedItem? → Need seed data tasks - Shared fixtures: What data appears across multiple stories?
Present the global setup plan to the developer.
Phase 2b: Implementation Order
Recommend an implementation order:
- Global setup (auth, DB reset, seed data)
- MVP stories, ordered by dependency (foundational flows first)
- SECONDARY stories
- ADVANCED stories
If focusing on a specific story:
Read the story's YAML, identify:
- What given/when/then steps are involved
- What systems are exercised
- What setup is needed
- Any custom DSL steps that need implementing
Phase 3: Iterative Implementation
Work through stories one at a time:
For each story:
- Read the YAML — understand the acceptance criteria
- Check prerequisites — is global setup done? Are dependent stories implemented?
- Decide approach using the decision tree:
- Does the DSL step need extending? → Add step to
cypress/support/steps.ts+ schema - Does the app need changes? → Implement the app feature
- Does test data need creating? → Add fixture or seed task
- Does the DSL step need extending? → Add step to
- Implement — write the code to make the story pass
- Verify — run the Cypress test for this story
- Commit — after the test passes, commit
Present each story to the developer before implementing:
Story: [story name]
Priority: [MVP/SECONDARY/ADVANCED]
Journey: [journey name]
Systems: [involved systems]
Steps:
given: [setup steps]
when: [action steps]
then: [assertion steps]
Implementation plan:
1. [what needs to be built]
2. [what needs to be configured]
3. [what tests to run]
Wait for developer approval before implementing each story.
Key Patterns
Auth setup: Create loginAs task in cypress.config.ts that sets localStorage/cookies for the test user.
DB reset: Add beforeEach in cypress/support/e2e.ts that calls an API endpoint or Cypress task to reset the database.
Seed data: Create seedItem task that inserts test data via API or direct DB access.
DSL extension: When a story uses a step not in the core DSL, add it to cypress/support/steps.ts and register in cypress/support/schema.ts.
Phase 4: Progress & Next Steps
After each story is implemented:
- Report progress: "X of Y MVP stories complete"
- Suggest the next story to implement
- Note any issues discovered during implementation
After all MVP stories pass:
- "MVP implementation complete. Move to SECONDARY stories?"
- Suggest running
/rs-validateto check coverage