docs: update documentation for pipeline changes (#218 Phase 3)

[let-sunny]

Summary

• CLAUDE.md: Update calibration flow description — Converter no longer runs visual-compare; orchestrator handles all measurements via CLI
• PROMPT.md: Replace "Rule Discovery A/B Validation" with "Ablation experiments" in pipeline list
• README.md: Remove "6-agent" reference from calibration description

Changes (3 files, 5 lines each)

All doc-only changes reflecting #220 (Rule Discovery removal), #221 (CLI unification), #222 (Converter role separation).

Part of #218 (Phase 3: documentation)

https://claude.ai/code/session_01N72z2Wbib4cLhYc3FRdSQi

Summary by CodeRabbit

• Documentation
  • Clarified calibration pipeline description, separating conversion (HTML generation) from measurement steps
  • Replaced specific tool reference with a unified “Measurements” stage that covers visual comparison, code metrics, and responsive checks
  • Removed explicit “6-agent calibration” wording and refined responsibility descriptions for clearer, user-facing expectations

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro

Run ID: 4d2407b7-2d84-4f01-b7b5-ca6d331eb382

📥 Commits

Reviewing files that changed from the base of the PR and between 24dbcd1 and f39cb64.

📒 Files selected for processing (2)

• .claude/commands/calibrate-loop.md
• src/agents/orchestrator.ts

---

📝 Walkthrough

Walkthrough

Documentation updates: pipeline target in PROMPT.md changed to "Ablation experiments (API-based)"; calibration docs (CLAUDE.md, calibrate-loop.md, README.md) reassign measurement responsibilities from the Converter to a new Measurements stage; small comment tweak in orchestrator.ts.

Changes

Cohort / File(s)	Summary
Prompt ./.claude/skills/design-to-code/PROMPT.md	Replaced pipeline target "Rule Discovery A/B Validation" with "Ablation experiments (API-based)".
Calibration docs CLAUDE.md, README.md, .claude/commands/calibrate-loop.md	Clarified pipeline: Converter narrowed to HTML generation; introduced Measurements stage (html-postprocess, visual-compare, code-metrics); updated tier descriptions and removed explicit "6-agent calibration pipeline" wording.
Agent comments src/agents/orchestrator.ts	Comment updates describing calibration tiers to reference "Measurements" instead of "visual-compare" (no logic changes).

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• refactor: separate Converter role + rename run-phase1 (#218 Phase 2) #222: Refactors Converter responsibilities and moves measurement steps to an orchestrator/Measurements stage (code-level counterpart).
• docs: update calibration and rule discovery documentation #88: Documentation realignment of the calibration pipeline and responsibilities between Converter and Measurements.
• feat: rule-discovery prompt fixes + deterministic CLI extraction (#145) #173: Changes to the design-to-code prompt surface that relate to PROMPT.md pipeline target updates.

Poem

🐰 A little rabbit hops through docs so bright,

Ablations bloom and measurements take flight,
The Converter trims to tidy HTML,
While Metrics dance and comparisons tell,
Hooray for clearer pipelines in sight!

🚥 Pre-merge checks | ✅ 2

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: documentation updates for pipeline architecture modifications, with specific reference to the associated phase.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch claude/issue-218-phase3-docs

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}