analyzthis_design 1.7.0 → 1.10.0

package/README.md

@@ -1,6 +1,6 @@
 # Analyzthis_Design
 
-A set of AI design personas and a task-first evaluation framework that plugs into Cursor, Claude Code, and Codex CLI as slash commands.
+A set of AI design personas and a task-first evaluation framework that plugs into Cursor, Claude Code, and Codex CLI as slash commands — plus an agentic MoE router with shared session state so you can call the same graph from any IDE or from the CLI.
 
 Install once. Run structured UX critiques, multi-phase ideation, and task-grounded screen reviews — directly inside your AI chat.
 
@@ -38,150 +38,289 @@ npx analyzthis_design --target all
 
 | Command | What it does |
 |---|---|
-| `/ux-story-gate` | **Recommended first step.** Discovers PRDs and user stories from your knowledge bank and repo, builds a task map, routes to the right personas, and synthesises findings by task with P0/P1/P2 priorities. |
-| `/design-critic` | 4-persona critique → `SHIP / REVISE / BLOCK` verdict with a Composite Score out of 20. |
+| `/persona-orchestrator` | **Recommended agentic entry point.** Loads MoE router + session state, runs ux-story-gate intake, executes the right persona chain, enforces DS / hierarchy / verify gates, synthesises a SHIP/REVISE/BLOCK verdict. |
+| `/ux-story-gate` | Task-first gate: discovers PRDs, DS/Figma discovery, MoE routing, browser verify, assess-only mode. |
+| `/design-critic` | 4-persona critique → `SHIP / REVISE / BLOCK` verdict with a Composite Score out of 20 + Information Hierarchy Gate. |
 | `/ux-ideator` | 6-phase ideation → two competing IA concepts, deliberation, delight pass, feasibility check. |
 
 ### 7 Individual Personas
 
-Invoke directly for targeted, already-grounded questions. For full screen evaluation, prefer `/ux-story-gate`.
+Invoke directly for targeted, already-grounded questions. For full screen evaluation, prefer `/persona-orchestrator` or `/ux-story-gate`.
 
 | Command | Persona | What they evaluate |
 |---|---|---|
-| `/arjun` | UX Agent | UX Honeycomb (7 dimensions, A–F): Useful, Usable, Findable, Credible, Accessible, Desirable, Valuable |
-| `/meera` | Business Agent | Retention, ARR, GTM lever, adoption risk by customer segment |
-| `/priya` | Feasibility Agent | Engineering effort (T-shirt sizing, 2-axis model), state machine traps, implementation risks |
-| `/zara` | Delight Agent | Picks exactly ONE peak delight moment — or says "speed is the craft" for working surfaces |
-| `/noor` | IA Architect | Minimalist, progressive-disclosure wireframe — Concept A, ≤3 nav levels |
-| `/anuj` | Power-User Advocate | Dense, expert-optimized wireframe — Concept B, bulk actions, keyboard shortcuts |
+| `/arjun` | UX + Visual Design | UX Honeycomb + Visual Design Audit (hierarchy, color, type, spacing, components, style fit, micro-interactions) |
+| `/meera` | Business Agent | Retention, ARR, GTM lever, adoption risk; hierarchy vs north-star check |
+| `/priya` | Feasibility Agent | Engineering effort (T-shirt sizing, 2-axis model), state machine traps |
+| `/zara` | Delight Agent | Exactly ONE peak delight moment — never contrast/token recovery (routes to DS Gate + Arjun) |
+| `/noor` | IA Architect | Minimalist Concept A + declared ranked information hierarchy |
+| `/anuj` | Power-User Advocate | Dense Concept B — bulk actions, keyboard shortcuts, hierarchy kept prominent |
 | `/raj` | Arbitrator | Resolves persona stalemates using 5 ranked product principles. Never speaks first. |
 
 ### Supporting skills
 
 | Command | Purpose |
 |---|---|
-| `/design-personas` | Session context template — fill in once before a session to ground all 7 personas in project-specific data |
+| `/design-personas` | Session context template — fill in once before a session |
 | `/knowledge-bank` | Auto-populated from your connected vault. All personas read this first. |
+| `/design-reference` | CSV reference data (colors, typography, UX guidelines, stacks, …) |
 
 ---
 
-## UX Story Gate — How it works
+## Agentic system (v1.10)
 
-`/ux-story-gate` is the task-first entry point for any screen evaluation. It runs in 5 phases:
+```
+User ask / Figma URL
+        ↓
+  /persona-orchestrator
+        ↓
+  ux-story-gate Phases 0–1.5 (PRD + DS/Figma + MoE router)
+        ↓
+  MoE subset (default, 1–2 experts) OR design-critic / ideation chain (explicit "full")
+        ↓
+  Hard gates: DS tokens → Information Hierarchy → Browser verify (skipped if assess_only + no URL)
+        ↓
+  Session state + cost metrics persisted  →  SHIP / REVISE / BLOCK
+```
+
+**Shared session state** lives at `~/.analyzthis_design/sessions/{project-id}/session-state.json` so Ask→Agent turns do not re-derive the task map, DS checklist, or routing decision.
 
-**Phase 0 — PRD Discovery (automatic)**
-Before asking you anything, it scans:
-1. Your connected knowledge bank for user stories, PRDs, acceptance criteria
-2. The current repo for `PRD*.md`, `docs/**/*.md`, `requirements/*.md`, and any file containing "user story / acceptance criteria / done when"
+```bash
+npx analyzthis_design session init
+npx analyzthis_design session show
+npx analyzthis_design session reset
+```
 
-It builds a draft task map from what it finds and asks you to confirm — you correct, not fill in from scratch.
+**Portable agent graph** (same manifests for Cursor / Claude / Codex / CLI):
 
-**Phase 1 — Task Map Gate**
-If no PRD context is found anywhere, it asks for the task map manually:
-- Which persona? (e.g. Retailer Admin, Media Sales)
-- What task? (verb + object)
-- How often? (daily / weekly / one-time)
-- Done when?
-- Fails when?
+```
+agents/
+  manifests/     # one JSON per persona + orchestrator
+  router.json    # MoE problem-type → expert list
+  chain.json     # default + ideation sequential graphs
+  session-schema.json
+```
+
+**Standalone runtime (v2):**
+
+```bash
+# Print routing only (no API calls)
+npx analyzthis_design run --task "Fix contrast on landing page" --dry-run
+
+# Call Anthropic / OpenAI per persona step (MoE subset, lite schema — the default)
+export ANTHROPIC_API_KEY=sk-...
+npx analyzthis_design run --task "Review this screen" --figma https://figma.com/... --provider anthropic
+
+# Force the full design-critic chain, or bypass the router entirely
+npx analyzthis_design run --task "Full critique of onboarding" --full
+npx analyzthis_design run --task "Just check spacing" --experts arjun
+```
+
+Provider defaults live in `~/.analyzthis_design/config.json`:
+
+```json
+{
+  "orchestrator": {
+    "provider": "anthropic",
+    "model": "claude-sonnet-5",
+    "mode": "lite",
+    "tiers": {
+      "structured": { "provider": "openai", "model": "gpt-4o-mini" },
+      "critique":   { "provider": "anthropic", "model": "claude-sonnet-5" },
+      "arbitrate":  { "provider": "anthropic", "model": "claude-sonnet-5" }
+    },
+    "max_tokens": { "structured": 900, "critique": 1800, "arbitrate": 1200 }
+  },
+  "pricing": {
+    "glm-4.5-flash":     { "input_per_m": 0,    "output_per_m": 0 },
+    "gemini-2.5-flash":  { "input_per_m": 0.30, "output_per_m": 2.50 },
+    "claude-sonnet-5":   { "input_per_m": 2,    "output_per_m": 10 },
+    "gpt-4o":            { "input_per_m": 2.50, "output_per_m": 10 }
+  },
+  "research": { "provider": "https://example.com/search?q={query}" }
+}
+```
 
-**Phase 2 — Field Veto Pass**
-Every field, button, and section on the screen gets a verdict:
-- ✅ Keep — maps to a task
-- ❌ Cut — no task owner
-- ⚠️ Clarify — ambiguous
+The `effort_matrix` and `gate_override` live in `agents/chain.json` (not the user config) so they ship with the package and stay in sync with the agent graph. `pricing` is user-configured so you control your own $-cost reporting.
 
-**Phase 3 — Scale & States**
-Forces declaration of scale (rows/entities), and checks for Empty / Error / Loading / Edge states before personas run.
+**Web research:**
 
-**Phase 4 — Persona Routing**
-Routes to the right personas based on task characteristics — not screen type.
+```bash
+npx analyzthis_design research --url https://example.com/design-tokens
+npx analyzthis_design research --query "EY design system tokens"
+```
 
-**Phase 5 — Synthesis**
-A Task × Finding table with P0 / P1 / P2 priorities and a build-ready verdict.
+Writes to `~/.analyzthis_design/sessions/{id}/web-context.md` and merges into the knowledge bank on `sync`.
 
 ---
 
-## Knowledge Bank — Connect your vault
+## Efficiency & cost (v1.10)
+
+The orchestrator defaults to the cheapest path that still respects every gate — fewer expert calls, shorter prompts, cheaper models where judgment isn't required, and on-disk caching. These savings apply to the **critique/audit** path (what this package does); see *What this actually saves* below for the honest scope.
+
+### Effort-graded model selection (v1.10)
+
+Each persona call is classified **trivial | standard | hard** from cheap signals already in the routing + session digest (no LLM call — a model call to pick a model would eat the savings). The classifier then resolves the model from an effort matrix, with persona-level overrides winning and the legacy `tiers` map as the final fallback so existing manifests keep working unchanged.
+
+```mermaid
+flowchart TB
+    Ask[User ask] --> Router[MoE router + effort classifier]
+    Router -->|effort| Resolve[resolveModel persona effort]
+    Resolve -->|gate? hard override| Matrix[effort_matrix in chain.json]
+    Resolve -->|persona| Overrides[manifest.effort_overrides]
+    Matrix --> Call[callLlm provider model maxTokens]
+    Overrides --> Call
+    Call --> Metrics[metrics.effort_log + cost_usd]
+    Metrics --> CostCmd[npx analyzthis_design cost]
+```
+
+Classifier rules (first match wins, safety rules before savings rules):
+- scoped mode active → **trivial** (single dimension by construction)
+- stalemate / any BLOCK / `full_chain` / `full_screen_review` → **hard**
+- `digest.ds_at_risk` non-empty → **hard**
+- REVISE delta follow-up → **trivial**
+- `manifest.tier == structured` → **trivial**, `arbitrate` → **standard**
+- default → **standard**
+
+**Gates never downgrade.** `ds_gate`, `information_hierarchy_gate`, and `verify_gate` are pinned to `hard` via `chain.gate_override` regardless of the classified effort — they're the safety net that makes downgrading persona work safe.
+
+Default effort matrix (in `agents/chain.json`):
+- trivial → `glm-4.5-flash` (free) or Gemini Flash-Lite, ~500-token cap
+- standard → `gemini-2.5-flash` or `gpt-4o-mini`, ~1200-token cap
+- hard → `claude-sonnet-5` or `gpt-5`, ~1800-token cap
 
-The knowledge bank lets you connect an Obsidian vault or any markdown folder so that personas read your actual project context — brand guidelines, design decisions, PRDs, research — before forming any opinion.
+Per-persona `effort_overrides` in each manifest refine this (e.g. Arjun's `trivial` is the color-system-only scoped mode at 700 tokens; his `hard` is the full Honeycomb + Visual Audit at 1800).
+
+```
+Ask → session digest → MoE router (1–2 experts, not 4) → persona cards (not full skills)
+    → retrieve-on-demand CSV rows (not whole files) → model tier by step → caches → cost metrics
+```
+
+| Lever | Default behavior |
+|---|---|
+| **Expert budget** | 1–2 personas per ask. Full `design-critic` chain only runs for an explicit "full critique" or `full_screen_review`. |
+| **Early DS exit** | Any "at risk" DS Token Checklist item stops the chain at `arjun_color_system_only` — Meera/Priya/Zara wait until it clears. |
+| **Delta re-evaluation** | A follow-up after REVISE re-runs only the personas assigned to the prior Top 3 changes, never the full chain. |
+| **Persona cards** | `agents/cards/<persona>.md` (~500 tokens) are the default system prompt; the full `skills/<persona>/SKILL.md` is only opened for a C-or-below rubric lookup or an explicit deep/full request. |
+| **Lite output schema** | Grades + Top 2 fixes + score, by default. Deep/full schema is opt-in. |
+| **Retrieve-on-demand** | `npx analyzthis_design retrieve --file colors.csv --column "Product Type" --keywords saas` returns only matching rows, pre-formatted for citation — never the whole CSV. |
+| **Model tiers** | `structured` steps can run on a cheaper model (e.g. `gpt-4o-mini`); `critique`/`arbitrate` steps use a stronger model. Configurable per tier in `~/.analyzthis_design/config.json`. |
+| **Caching** | `lib/cache.js` caches retrieve results (invalidated automatically when the source CSV changes) and knowledge-bank slices (invalidated on `sync` / `session reset`). |
+| **Cost metrics** | Every `run` records `metrics` (llm_calls, experts_run, estimated tokens, cache_hits) into session state. |
 
 ```bash
-# Connect a vault (all notes)
-npx analyzthis_design connect --vault ~/Documents/MyVault
+npx analyzthis_design metrics                 # last run's cost summary for this project
+npx analyzthis_design metrics --all           # across every project
+```
 
-# Connect with tag filter (only notes tagged #design, #brand, #prd)
-npx analyzthis_design connect --vault ~/vault --tags design,brand,prd,product
+### What this actually saves (and what it doesn't)
 
-# Connect with folder filter
-npx analyzthis_design connect --vault ~/vault --include Design,Brand,PRDs,Research
+analyzthis_design is a design **critique** layer, not a design generator. The personas *review* UI; they don't produce a finished design end-to-end. So the savings show up on the **review** side of the loop, and across the **create → review → revise** loop when your host LLM uses the personas as a guided check — not on raw generation in isolation.
 
-# Sync to Cursor
-npx analyzthis_design sync
+**Honest, measurable savings on the critique path:**
 
-# Sync to all tools
-npx analyzthis_design sync --target all
+- ~50–75% fewer expert LLM calls on narrow asks (1–2 personas vs. 4).
+- ~50%+ fewer input tokens per `run` (persona cards vs. full SKILL.md).
+- Retrieve-on-demand sends only matching CSV rows, not whole files (`colors.csv` is 32 kB, `styles.csv` is 143 kB — we send ~5 rows).
+- Structured/extract steps can run on a cheaper model with a 900-token cap; only critique/arbitrate uses the strong model.
+- Repeat runs on the same file hit the cache instead of re-processing Figma screenshots, KB slices, and CSV packs.
+- Every saving above is **observable** via `npx analyzthis_design metrics` (`llm_calls`, `input_tokens_est`, `output_tokens_est`, `cache_hits`).
 
-# Check what's connected
-npx analyzthis_design status
+**Where the savings come from across the whole loop** (when the host LLM routes a design through the personas):
 
-# Remove a source
-npx analyzthis_design disconnect --vault ~/Documents/MyVault
+- Fewer revision rounds — DS / hierarchy / contrast failures are caught early instead of after a full review.
+- Data-driven citations ground the LLM so it doesn't hallucinate or re-derive design rules.
+- The host LLM gets a compact digest + targeted fixes, not a wall of prose.
+
+**What this is *not*:**
+
+- It does **not** generate end-to-end designs using fewer tokens — it critiques.
+- It does **not** save tokens vs. "using no AI at all" — it adds a review layer; it saves tokens vs. an *unstructured* review loop.
+- There is no hard percentage claim yet — v1.9 ships *targets* (full-chain rate <30%, median experts ≤2, ~50% fewer skill-prompt tokens), not proven production numbers. Run `metrics` on your own workload to see your actual savings.
+
+**LoRA readiness (export hook only — no training in this release):**
+
+```bash
+npx analyzthis_design session accept --persona arjun     # mark the last output as a good example
+npx analyzthis_design export-training --persona arjun --all
 ```
 
-Once synced, the knowledge bank is auto-read at the start of every persona session. Your project-specific context takes full precedence over built-in persona defaults.
+Writes `{ system_card, digest, user, assistant }` JSONL pairs to `~/.analyzthis_design/training/<persona>.jsonl` from every session where that persona's output was explicitly accepted. Once a persona accumulates ~100–300 accepted pairs, that data is ready for a future fine-tuning pass on an open model — not part of this package yet.
 
-**PRDs and user stories** in your vault are automatically detected and surfaced at the top of the knowledge bank — `ux-story-gate` reads them in Phase 0 to build the task map without you having to type it out.
+---
+
+## UX Story Gate — How it works
 
-Config is stored at `~/.analyzthis_design/config.json` — global across all projects.
+`/ux-story-gate` is the task-first gate for any screen evaluation:
+
+| Phase | What it does |
+|---|---|
+| 0 | PRD discovery from knowledge bank + repo |
+| 0.5 | DS / Figma discovery + DS Token Checklist (exit criteria) |
+| 1 | Task map intake gate |
+| 1.5 | MoE problem-type router → writes `routing_decision` to session state |
+| 2 | Field veto pass |
+| 3 | Scale & states declaration |
+| 4 | Per-task persona routing |
+| 4.5 | Browser verify gate (navigate → snapshot → primary flow → mobile+desktop screenshot) |
+| 5 | Task × Finding synthesis |
+| 5.5 | Assess-only mode — no code changes until you say build / implement / apply |
 
 ---
 
-## CLI Reference
+## Knowledge Bank — Connect your vault
 
 ```bash
-# ── Install ────────────────────────────────────────────────────────────────
-npx analyzthis_design                          # install for Cursor
-npx analyzthis_design --target claude          # install for Claude Code
-npx analyzthis_design --target codex           # install for Codex CLI
-npx analyzthis_design --target all             # install for all tools
-npx analyzthis_design --force                  # overwrite existing skills
-
-# ── Remove ─────────────────────────────────────────────────────────────────
-npx analyzthis_design remove                   # remove from Cursor
-npx analyzthis_design remove --target all      # remove from all tools
-
-# ── List ───────────────────────────────────────────────────────────────────
-npx analyzthis_design list                     # show Cursor installs
-npx analyzthis_design list --target all        # show all tool installs
-
-# ── Knowledge bank ─────────────────────────────────────────────────────────
-npx analyzthis_design connect --vault <path>
-npx analyzthis_design connect --vault <path> --tags design,prd,brand
-npx analyzthis_design connect --vault <path> --include Design,PRDs,Research
+npx analyzthis_design connect --vault ~/Documents/MyVault
+npx analyzthis_design connect --vault ~/vault --tags design,brand,prd,product
+npx analyzthis_design connect --vault ~/vault --include Design,Brand,PRDs,Research
 npx analyzthis_design sync
 npx analyzthis_design sync --target all
-npx analyzthis_design disconnect --vault <path>
 npx analyzthis_design status
+npx analyzthis_design disconnect --vault ~/Documents/MyVault
 ```
 
+PRDs and user stories are surfaced at the top of the knowledge bank. Brand / design-system notes feed Phase 0.5. Web research merges under **Web Research Context**.
+
+Config: `~/.analyzthis_design/config.json`.
+
 ---
 
-## How it fits together
+## CLI Reference
 
-```
-Your vault / repo PRDs
-        ↓
-  knowledge bank  ←── npx analyzthis_design sync
-        ↓
-  /ux-story-gate
-        ↓
-  Phase 0: read knowledge bank + scan repo for PRDs
-        ↓
-  Build task map (or confirm draft from PRDs)
-        ↓
-  Route to: /noor  /anuj  /arjun  (+/meera  /zara  as needed)
-        ↓
-  Task × Finding table  →  P0 / P1 / P2 verdict
+```bash
+# Install / remove / list
+npx analyzthis_design
+npx analyzthis_design --target all
+npx analyzthis_design --force
+npx analyzthis_design remove --target all
+npx analyzthis_design list --target all
+
+# Knowledge bank
+npx analyzthis_design connect --vault <path> [--tags ...] [--include ...]
+npx analyzthis_design sync [--target all]
+npx analyzthis_design disconnect --vault <path>
+npx analyzthis_design status
+
+# Session (agentic)
+npx analyzthis_design session init|show|reset [--project id] [--all]
+npx analyzthis_design session accept --persona <id> [--reject]
+
+# Research
+npx analyzthis_design research --url <url>
+npx analyzthis_design research --query <text>
+
+# Reference data (retrieve-on-demand)
+npx analyzthis_design retrieve --file <csv> --column <col> --keywords a,b [--limit N]
+
+# Standalone orchestrator
+npx analyzthis_design run --task "..." [--figma URL] [--provider anthropic|openai] [--dry-run] [--output path]
+npx analyzthis_design run --task "..." [--lite | --full] [--experts a,b]
+
+# Efficiency / cost
+npx analyzthis_design metrics [--project id] [--all]
+npx analyzthis_design cost [--project id] [--all]
+npx analyzthis_design export-training --persona <id> [--project id] [--all] [--output path]
 ```
 
 ---
@@ -189,28 +328,27 @@ Your vault / repo PRDs
 ## Repository structure
 
 ```
-bin/
-  cli.js              CLI entry point
+agents/                 Portable MoE graph (manifests, router, chain, session schema)
+  cards/                Short per-persona system prompts (~500 tokens each)
+bin/cli.js              CLI entry point
 lib/
-  install.js          Skill installation logic (copies to ~/.cursor/skills/ etc.)
-  knowledge.js        Vault reading, sync, PRD categorisation
-scripts/
-  obfuscate.js        Build step — obfuscates lib/ and bin/ into dist/ before publish
+  install.js            Skill installation
+  knowledge.js          Vault sync + web-context merge
+  session.js            Shared session-state.json (+ digest, metrics, accept flag)
+  research.js           URL / query → web-context.md
+  retrieve.js           Filtered, citation-ready CSV row retrieval
+  cache.js              On-disk cache for retrieve/kb slices
+  export.js             LoRA training-pair export hook
+  cost.js               $-cost report from metrics × config.pricing
+  orchestrator/run.js   Standalone LLM runtime (v2) — MoE, effort-graded tiers, caps, cache-aware
+scripts/obfuscate.js    Build step → dist/
 skills/
-  arjun/              UX Agent — Honeycomb scoring
-  meera/              Business Agent — retention, ARR, GTM
-  priya/              Feasibility Agent — effort sizing
-  zara/               Delight Agent — peak-end moment
-  noor/               IA Architect — minimalist Concept A
-  anuj/               Power-User Advocate — dense Concept B
-  raj/                Arbitrator — stalemate resolution
-  design-critic/      Orchestrator — 4-persona critique
-  ux-ideator/         Orchestrator — 6-phase ideation
-  ux-story-gate/      Orchestrator — task-first gate + PRD discovery
-  design-personas/    Session context template
-  knowledge-bank/     Auto-populated from connected vault
-  design-reference/   12 CSV files — colors, typography, UX guidelines,
-                      component patterns, charts, icons, 16 framework stacks
+  persona-orchestrator/ Agentic entry point
+  ux-story-gate/        Task-first gate + DS/MoE/verify/assess phases
+  design-critic/        4-persona critique + hierarchy gate
+  ux-ideator/           6-phase ideation
+  arjun/ meera/ priya/ zara/ noor/ anuj/ raj/
+  design-personas/ knowledge-bank/ design-reference/
 ```
 
 ---
@@ -220,6 +358,7 @@ skills/
 - Node.js 16+
 - [Cursor](https://cursor.com) with Agent Mode (for `/skill` commands)
 - Claude Code or Codex CLI if using those targets
+- For `run` (non–dry-run): `ANTHROPIC_API_KEY` or `OPENAI_API_KEY`
 
 ---
 

package/agents/cards/anuj.md

@@ -0,0 +1,22 @@
+# Anuj (Dev) — Power User (card)
+
+6 years domain analyst in high-volume ops before product. Information density is a feature, not a flaw.
+
+**Allowed:** audit density, bulk actions, keyboard shortcuts for daily-use surfaces; produce Concept B wireframe; keep rank #1 of the declared information hierarchy prominent even at full density.
+
+**Forbidden:** proposing density audits for surfaces with task Frequency ≠ daily/weekly; code edits without explicit build approval.
+
+**Non-negotiable:** density never flattens the information hierarchy — rank #1 stays the most prominent element (leftmost/largest/first-sorted) even at full density.
+
+## Lite output schema (default)
+```
+## Anuj — Lite
+Hierarchy check: rank #1 [element] kept prominent via [mechanism]
+Bulk actions: [present/missing] — Keyboard shortcuts: [list or "none"]
+Top fix: [one line]
+```
+
+## Deep output schema
+Full Concept B text wireframe (hierarchy check, primary action, nav level, visible-on-load at full density, keyboard shortcuts, progressive disclosure, rationale) — see `skills/anuj/SKILL.md` "Output — Concept B". Use deep mode for ideation runs or full/deep critiques.
+
+**Citation:** `[filename, row N: "exact quoted value"]` — e.g. `[ux-guidelines.csv, row 22: "..."]`. Only run this persona when task Frequency = daily/weekly.

package/agents/cards/arjun.md

@@ -0,0 +1,23 @@
+# Arjun — UX + Visual Design (card)
+
+Product designer (200+ user sessions, B2B SaaS) turned design-systems lead (3 yrs, 200+ shipped components). Runs both the UX lens and the visual-design lens in one pass.
+
+**Allowed:** UX Honeycomb critique (Useful, Usable, Findable, Credible, Accessible, Desirable, Valuable); full Visual Design Audit (hierarchy, color, typography, spacing, components, style fit, micro-interactions); diagnosing visual issues against declared information hierarchy + DS tokens.
+
+**Forbidden:** brand-system recovery as a primary job (diagnostic only, no `!important` patches); delight pass (→ Zara); code edits without explicit build approval.
+
+**Scoped mode `arjun_color_system_only`:** used on a DS Gate early exit — grade only Color System + Typography contrast, skip the rest of the Honeycomb/Visual Audit.
+
+## Lite output schema (default)
+```
+## Arjun — Lite
+UX grades: Useful[A-F] Usable[A-F] Findable[A-F] Credible[A-F] Accessible[A-F] Desirable[A-F] Valuable[A-F]
+Visual grades: Hierarchy[A-F] Color[A-F] Type[A-F] Spacing[A-F] Components[A-F] StyleFit[A-F] Micro[A-F]
+Top 2 fixes: 1. [component+zone+fix] 2. [component+zone+fix]
+Combined score: [X/5]
+```
+
+## Deep output schema
+Full UX Critique + Visual Design Audit blocks with per-dimension reasons — see `skills/arjun/SKILL.md` "Output format". Use deep mode when the user asks for a full/deep critique, any dimension scores C or below and needs the full rubric quoted, or `default_chain` is running.
+
+**Citation:** `[filename, row N: "exact quoted value"]`. Consult `skills/arjun/SKILL.md` Grade Rubric tables when scoring C or below and rubric detail is needed.

package/agents/cards/meera.md

@@ -0,0 +1,21 @@
+# Meera — Business (card)
+
+Ex-revenue/sales. Thinks in retention, ARR, GTM levers. Numbers-first, segmentation-aware. Skeptical of features that demo well but die in production adoption.
+
+**Allowed:** north-star metric impact assessment; segment/GTM/retention analysis; check that rank #1 on screen matches the actual business-critical driver.
+
+**Forbidden:** visual or UX critique (→ Arjun); code edits without explicit build approval.
+
+## Lite output schema (default)
+```
+## Meera — Lite
+Metric impact: [moves it / neutral / hurts it]
+Hierarchy check: [matches / does not match] north-star driver
+Top 2 fixes: 1. [...] 2. [...]
+Score: [1-5]
+```
+
+## Deep output schema
+Full Business Impact block (metric impact, hierarchy check, segment, GTM lever, retention hook, adoption risk, verdict) — see `skills/meera/SKILL.md` "Output format". Use deep mode for full/deep critiques or when `default_chain` is running.
+
+**Citation:** `[filename, row N: "exact quoted value"]` — e.g. `[products.csv, row 6: "..."]`. Consult `skills/meera/SKILL.md` for the full reference-data table when a specific product-type match is needed.

package/agents/cards/noor.md

@@ -0,0 +1,22 @@
+# Noor — Minimalist IA (card)
+
+7 years IA for SaaS across fintech, workflow automation, B2B tooling. Shipped at 50k–500k DAU — scale punishes complexity, doesn't justify it.
+
+**Allowed:** declare ranked information hierarchy; propose minimalist IA / progressive-disclosure structure; produce Concept A wireframe.
+
+**Forbidden:** brand token recovery; contrast/accessibility fixes (→ Arjun); code edits without explicit build approval.
+
+**Non-negotiable:** information hierarchy is declared before any layout decision — this ranking is ground truth for Anuj (density), Meera (business-critical info), and Arjun (visual weight).
+
+## Lite output schema (default)
+```
+## Noor — Lite
+Hierarchy (ranked): 1. [...] 2. [...] 3. [...]
+Primary action: [CTA] — Nav level: L[1/2/3]
+Rationale: [one line, Hick's Law / progressive disclosure]
+```
+
+## Deep output schema
+Full Concept A text wireframe (hierarchy, primary action, nav level, visible-on-load, progressive disclosure, nav path, rationale) — see `skills/noor/SKILL.md` "Output — Concept A". Use deep mode for ideation runs or full/deep critiques.
+
+**Citation:** `[filename, row N: "exact quoted value"]` — e.g. `[stacks/shadcn.csv, row 8: "..."]`. Consult `skills/noor/SKILL.md` when naming specific components.

package/agents/cards/priya.md

@@ -0,0 +1,21 @@
+# Priya — Feasibility (card)
+
+Senior full-stack engineer, 8+ years complex SaaS. Blunt, precise. Has been burned by "simple UI change" features that became 3-month infra projects.
+
+**Allowed:** T-shirt sizing (two-axis: UI × State); risk/blocker identification; simpler-alternative sizing.
+
+**Forbidden:** visual or business critique; code edits without explicit build approval.
+
+## Lite output schema (default)
+```
+## Priya — Lite
+Effort: [S/M/L/XL] — UI [S/M/L/XL] x State [S/M/L/XL]
+Top 2 risks: 1. [risk+consequence] 2. [risk+consequence]
+Simpler alternative: [one line or "none"]
+Score: [1-5]
+```
+
+## Deep output schema
+Full Feasibility Analysis block (score, blockers, ranked risks, effort, simpler alternative) — see `skills/priya/SKILL.md` "Output format". Use deep mode for full/deep critiques or when `default_chain` is running.
+
+**Citation:** `[filename, row N: "exact quoted value"]` — e.g. `[stacks/nextjs.csv, row 12: "..."]`. Consult `skills/priya/SKILL.md` for the stack-file table matching the session's tech stack.

package/agents/cards/raj.md

@@ -0,0 +1,23 @@
+# Raj — Arbitrator (card)
+
+10+ years product strategy. Speaks ONLY when the Stalemate Protocol activates — does not volunteer opinions. Every position anchored to PRD evidence, user data, or a named product principle.
+
+**Allowed:** resolve stalemates between personas using the 5 product principles; issue final SHIP/REVISE/BLOCK when personas disagree.
+
+**Forbidden:** running with no stalemate/BLOCK condition; code edits without explicit build approval.
+
+**Activation criteria (any one):** 2+ unconceded structural objections; a "non-negotiable" claim refused; the same argument repeated without new evidence; a PRD-persona-priority conflict with no established priority.
+
+## Output schema (always full — Raj has no lite mode; his output is inherently a short decision)
+```
+## Raj — Stalemate Resolution
+Activated by: [criterion]
+Contested dimensions: [...]
+PRD anchor: "[quote]"
+Product principle applied: [1-5, see skills/raj/SKILL.md]
+Decision: [resolution]
+Rationale: [2-3 sentences]
+What [losing agent] gives up: [named]
+```
+
+**Citation:** `[filename, row N: "exact quoted value"]`. Consult `skills/raj/SKILL.md` for the ranked product-principles list (Owner governs, Data honesty, Intentionality over automation, Persona density split, PRD scope boundary) and their worked examples before deciding.

package/agents/cards/zara.md

@@ -0,0 +1,24 @@
+# Zara — Delight (card)
+
+Consumer-app designer who brought the consumer-delight lens to B2B. The Peak-End Rule is her north star. Picks exactly ONE memorable moment over five forgettable ones.
+
+**Allowed:** identify exactly ONE structural or surface delight moment, on top of an already DS-compliant, hierarchy-correct foundation.
+
+**Forbidden:** contrast fixes, token drift, or any brand/DS recovery — refuse and route to DS Gate + Arjun; running before the DS Gate has passed; code edits without explicit build approval.
+
+**Gate check:** if session state `ds_checklist` has any item "at risk," refuse and re-route — do not produce a delight pass.
+
+## Lite output schema (default)
+```
+## Zara — Lite
+Moment: [where] — Type: [Structural/Surface]
+Specific addition: [one concrete detail]
+Cost: [low/medium/high]
+Score: [1-5]
+```
+If high-frequency working surface: output only "no delight needed here — speed is the craft."
+
+## Deep output schema
+Full Delight Pass block (surface, moment, type, specific addition, why-this-one, cost, design-system pointer) — see `skills/zara/SKILL.md` "Output format". Use deep mode for full/deep critiques or when `default_chain` is running.
+
+**Citation:** `[filename, row N: "exact quoted value"]`. Apply the `styles.csv` 5-step filter in `skills/zara/SKILL.md` before citing — never read the full file.