novahiz 0.1.0

package/bundled-skills/ads/references/platform-setup-checklists.md

@@ -0,0 +1,277 @@
+# Platform Setup Checklists
+
+Complete setup checklists for major ad platforms.
+
+## Contents
+- Google Ads Setup (Account Foundation, Conversion Tracking, Analytics Integration, Audience Setup, Campaign Readiness, Ad Extensions, Brand Protection)
+- Meta Ads Setup (Business Manager Foundation, Pixel & Tracking, Domain & Aggregated Events, Audience Setup, Catalog, Creative Assets, Compliance)
+- LinkedIn Ads Setup (Campaign Manager Foundation, Insight Tag & Tracking, Audience Setup, Lead Gen Forms, Document Ads, Creative Assets, Budget Considerations)
+- Twitter/X Ads Setup (Account Foundation, Tracking, Audience Setup, Creative)
+- TikTok Ads Setup (Account Foundation, Pixel & Tracking, Audience Setup, Creative)
+- Universal Pre-Launch Checklist
+
+## Google Ads Setup
+
+### Account Foundation
+
+- [ ] Google Ads account created and verified
+- [ ] Billing information added
+- [ ] Time zone and currency set correctly
+- [ ] Account access granted to team members
+
+### Conversion Tracking
+
+- [ ] Google tag installed on all pages
+- [ ] Conversion actions created (purchase, lead, signup)
+- [ ] Conversion values assigned (if applicable)
+- [ ] Enhanced conversions enabled
+- [ ] Test conversions firing correctly
+- [ ] Import conversions from GA4 (optional)
+
+### Analytics Integration
+
+- [ ] Google Analytics 4 linked
+- [ ] Auto-tagging enabled
+- [ ] GA4 audiences available in Google Ads
+- [ ] Cross-domain tracking set up (if multiple domains)
+
+### Audience Setup
+
+- [ ] Remarketing tag verified
+- [ ] Website visitor audiences created:
+  - All visitors (180 days)
+  - Key page visitors (pricing, demo, features)
+  - Converters (for exclusion)
+- [ ] Customer match lists uploaded
+- [ ] Similar audiences enabled
+
+### Campaign Readiness
+
+- [ ] Negative keyword lists created:
+  - Universal negatives (free, jobs, careers, reviews, complaints)
+  - Competitor negatives (if needed)
+  - Irrelevant industry terms
+- [ ] Location targeting set (include/exclude)
+- [ ] Language targeting set
+- [ ] Ad schedule configured (if B2B, business hours)
+- [ ] Device bid adjustments considered
+
+### Ad Extensions
+
+- [ ] Sitelinks (4-6 relevant pages)
+- [ ] Callouts (key benefits, offers)
+- [ ] Structured snippets (features, types, services)
+- [ ] Call extension (if phone leads valuable)
+- [ ] Lead form extension (if using)
+- [ ] Price extensions (if applicable)
+- [ ] Image extensions (where available)
+
+### Brand Protection
+
+- [ ] Brand campaign running (protect branded terms)
+- [ ] Competitor campaigns considered
+- [ ] Brand terms in negative lists for non-brand campaigns
+
+---
+
+## Meta Ads Setup
+
+### Business Manager Foundation
+
+- [ ] Business Manager created
+- [ ] Business verified (if running certain ad types)
+- [ ] Ad account created within Business Manager
+- [ ] Payment method added
+- [ ] Team access configured with proper roles
+
+### Pixel & Tracking
+
+- [ ] Meta Pixel installed on all pages
+- [ ] Standard events configured:
+  - PageView (automatic)
+  - ViewContent (product/feature pages)
+  - Lead (form submissions)
+  - Purchase (conversions)
+  - AddToCart (if e-commerce)
+  - InitiateCheckout (if e-commerce)
+- [ ] Conversions API (CAPI) set up for server-side tracking
+- [ ] Event Match Quality score > 6
+- [ ] Test events in Events Manager
+
+### Domain & Aggregated Events
+
+- [ ] Domain verified in Business Manager
+- [ ] Aggregated Event Measurement configured
+- [ ] Top 8 events prioritized in order of importance
+- [ ] Web events prioritized for iOS 14+ tracking
+
+### Audience Setup
+
+- [ ] Custom audiences created:
+  - Website visitors (all, 30/60/90/180 days)
+  - Key page visitors
+  - Video viewers (25%, 50%, 75%, 95%)
+  - Page/Instagram engagers
+  - Customer list uploaded
+- [ ] Lookalike audiences created (1%, 1-3%)
+- [ ] Saved audiences for common targeting
+
+### Catalog (E-commerce)
+
+- [ ] Product catalog connected
+- [ ] Product feed updating correctly
+- [ ] Catalog sales campaigns enabled
+- [ ] Dynamic product ads configured
+
+### Creative Assets
+
+- [ ] Images in correct sizes:
+  - Feed: 1080x1080 (1:1)
+  - Stories/Reels: 1080x1920 (9:16)
+  - Landscape: 1200x628 (1.91:1)
+- [ ] Videos in correct formats
+- [ ] Ad copy variations ready
+- [ ] UTM parameters in all destination URLs
+
+### Compliance
+
+- [ ] Special Ad Categories declared (if housing, credit, employment, politics)
+- [ ] Landing page complies with Meta policies
+- [ ] No prohibited content in ads
+
+---
+
+## LinkedIn Ads Setup
+
+### Campaign Manager Foundation
+
+- [ ] Campaign Manager account created
+- [ ] Company Page connected
+- [ ] Billing information added
+- [ ] Team access configured
+
+### Insight Tag & Tracking
+
+- [ ] LinkedIn Insight Tag installed on all pages
+- [ ] Tag verified and firing
+- [ ] Conversion tracking configured:
+  - URL-based conversions
+  - Event-specific conversions
+- [ ] Conversion values set (if applicable)
+
+### Audience Setup
+
+- [ ] Matched Audiences created:
+  - Website retargeting audiences
+  - Company list uploaded (for ABM)
+  - Contact list uploaded
+- [ ] Lookalike audiences created
+- [ ] Saved audiences for common targeting
+
+### Lead Gen Forms (if using)
+
+- [ ] Lead gen form templates created
+- [ ] Form fields selected (minimize for conversion)
+- [ ] Privacy policy URL added
+- [ ] Thank you message configured
+- [ ] CRM integration set up (or CSV export process)
+
+### Document Ads (if using)
+
+- [ ] Documents uploaded (PDF, PowerPoint)
+- [ ] Gating configured (full gate or preview)
+- [ ] Lead gen form connected
+
+### Creative Assets
+
+- [ ] Single image ads: 1200x627 (1.91:1) or 1080x1080 (1:1)
+- [ ] Carousel images ready
+- [ ] Video specs met (if using)
+- [ ] Ad copy within character limits:
+  - Intro text: 600 max, 150 recommended
+  - Headline: 200 max, 70 recommended
+
+### Budget Considerations
+
+- [ ] Budget realistic for LinkedIn CPCs ($8-15+ typical)
+- [ ] Audience size validated (50K+ recommended)
+- [ ] Daily vs. lifetime budget decided
+- [ ] Bid strategy selected
+
+---
+
+## Twitter/X Ads Setup
+
+### Account Foundation
+
+- [ ] Ads account created
+- [ ] Payment method added
+- [ ] Account verified (if required)
+
+### Tracking
+
+- [ ] Twitter Pixel installed
+- [ ] Conversion events created
+- [ ] Website tag verified
+
+### Audience Setup
+
+- [ ] Tailored audiences created:
+  - Website visitors
+  - Customer lists
+- [ ] Follower lookalikes identified
+- [ ] Interest and keyword targets researched
+
+### Creative
+
+- [ ] Tweet copy within 280 characters
+- [ ] Images: 1200x675 (1.91:1) or 1200x1200 (1:1)
+- [ ] Video specs met (if using)
+- [ ] Cards configured (website, app, etc.)
+
+---
+
+## TikTok Ads Setup
+
+### Account Foundation
+
+- [ ] TikTok Ads Manager account created
+- [ ] Business verification completed
+- [ ] Payment method added
+
+### Pixel & Tracking
+
+- [ ] TikTok Pixel installed
+- [ ] Events configured (ViewContent, Purchase, etc.)
+- [ ] Events API set up (recommended)
+
+### Audience Setup
+
+- [ ] Custom audiences created
+- [ ] Lookalike audiences created
+- [ ] Interest categories identified
+
+### Creative
+
+- [ ] Vertical video (9:16) ready
+- [ ] Native-feeling content (not too polished)
+- [ ] First 3 seconds are compelling hooks
+- [ ] Captions added (most watch without sound)
+- [ ] Music/sounds selected (licensed if needed)
+
+---
+
+## Universal Pre-Launch Checklist
+
+Before launching any campaign:
+
+- [ ] Conversion tracking tested with real conversion
+- [ ] Landing page loads fast (<3 sec)
+- [ ] Landing page mobile-friendly
+- [ ] UTM parameters working
+- [ ] Budget set correctly (daily vs. lifetime)
+- [ ] Start/end dates correct
+- [ ] Targeting matches intended audience
+- [ ] Ad creative approved
+- [ ] Team notified of launch
+- [ ] Reporting dashboard ready

package/bundled-skills/ads/references/rsa-output-spec.md

@@ -0,0 +1,88 @@
+# Google RSA Output Spec
+
+When the user requests Google Ads RSAs (Responsive Search Ads), output MUST comply with these platform limits and structural requirements. Do not output any RSA that violates them.
+
+## Hard limits per RSA (enforce before responding)
+
+- **Headlines:** exactly **15** per RSA, each **≤ 30 characters** (count characters, including spaces). Render as `1. ... (NN chars)` so the reader can verify.
+- **Descriptions:** exactly **4** per RSA, each **≤ 90 characters**.
+- **Paths:** up to 2 path fields, each **≤ 15 characters**.
+- **Final URL:** present, https.
+- **Pinning:** state any pinned positions explicitly. Default = unpinned unless user asks.
+- **Per-account guardrail:** Google enforces **3 RSAs max per ad group**. When the user asks for >3, group them by ad group.
+
+## Required sidecar artifacts (always include with RSA request)
+
+1. **Ad group structure**, labeled `Ad group structure:` — list each ad group with its theme, target keywords (match types), and which RSAs map to it.
+2. **Negative keyword list**, labeled `Negative keywords:` — minimum **8** entries, group-level vs campaign-level called out.
+3. **Sitelinks** (≥ 4), **Callouts** (≥ 4 ≤25 chars), **Structured snippets** if relevant.
+
+## Medical / CFM compliance (when product context indicates pt-BR medical practice)
+
+If `.agents/product-marketing.md` indicates a Brazilian medical practice (CFM-regulated), the following terms are **forbidden** in headlines, descriptions, sitelinks, and callouts:
+
+- Superlatives: `#1`, `melhor`, `o melhor`, `melhor do brasil`, `top`, `referência`
+- Outcome promises: `garantido`, `garantia`, `cura`, `cura definitiva`, `100%`, `resultado garantido`, `livre da dor`
+- Comparative claims vs other doctors/clinics
+
+Use neutral framing: `atendimento`, `consulta`, `avaliação`, `segunda opinião`, `agende sua consulta`, `tire suas dúvidas`. Geo modifier (`Porto Alegre`, `POA`, `Zona Sul POA`) required where the prompt specifies a region.
+
+## Output ORDER (mandatory — emit in this order to avoid truncation)
+
+1. **Ad group structure** (short)
+2. **Negative keywords** (≥8, MANDATORY — emit BEFORE RSAs so it isn't dropped if output runs long)
+3. **Sitelinks** (≥4)
+4. **Callouts** (≥4)
+5. **RSA1, RSA2, RSA3** (largest section, last — safe to truncate gracefully)
+
+## Output template (mandatory shape)
+
+```
+Ad group structure:
+- AG1 [theme]: keywords (match types) → RSA1, RSA2
+- AG2 [theme]: ...
+
+Negative keywords:
+  Campaign-level:
+    - <kw>
+    - <kw>
+    (≥4 here)
+  Ad-group level:
+    - AG1: <kw>, <kw>
+    - AG2: <kw>, <kw>
+    (≥4 more here — TOTAL ≥8 entries)
+
+Sitelinks (≥4):
+  - <title (≤25)> | <desc1 (≤35)> | <desc2 (≤35)> | URL
+
+Callouts (≥4, each ≤25 chars):
+  - <callout>
+
+RSA1 — [ad group name]
+  Final URL: https://...
+  Path1: ...   Path2: ...
+  Headlines (15, each ≤30 chars):
+    1. <headline> (NN chars)
+    ...
+    15. <headline> (NN chars)
+  Descriptions (4, each ≤90 chars):
+    1. <description> (NN chars)
+    ...
+    4. <description> (NN chars)
+  Pinning: H1=none; H2=none; ...   (or explicit pins)
+
+RSA2 — ...
+RSA3 — ...
+```
+
+## Self-check before responding
+
+Before sending the output, run this checklist mentally:
+
+- [ ] Each RSA has exactly 15 headlines, exactly 4 descriptions.
+- [ ] Every headline is ≤30 chars; every description is ≤90 chars. Character counts printed.
+- [ ] Negative keyword list labeled and ≥8 entries.
+- [ ] Ad group structure labeled.
+- [ ] If medical (CFM): no forbidden superlative/outcome words; geo modifier present where required; language is pt-BR.
+
+If any check fails, rewrite before responding. Do not ship partial RSAs.

package/bundled-skills/adversarial-reviewer/SKILL.md

@@ -0,0 +1,247 @@
+---
+name: "adversarial-reviewer"
+description: "Adversarial code review that breaks the self-review monoculture. Use when you want a genuinely critical review of recent changes, before merging a PR, or when you suspect Claude is being too agreeable about code quality. Forces perspective shifts through hostile reviewer personas that catch blind spots the author's mental model shares with the reviewer."
+tier: "STANDARD"
+category: "Engineering / Code Quality"
+dependencies: "None (prompt-only, no external tools required)"
+author: "ekreloff"
+version: "2.9.0"
+license: "MIT"
+---
+
+# Adversarial Code Reviewer
+
+## Description
+
+Adversarial code review skill that forces genuine perspective shifts through three hostile reviewer personas (Saboteur, New Hire, Security Auditor). Each persona MUST find at least one issue — no "LGTM" escapes. Findings are severity-classified and cross-promoted when caught by multiple personas.
+
+## Features
+
+- **Three adversarial personas** — Saboteur (production breaks), New Hire (maintainability), Security Auditor (OWASP-informed)
+- **Mandatory findings** — Each persona must surface at least one issue, eliminating rubber-stamp reviews
+- **Severity promotion** — Issues caught by 2+ personas are promoted one severity level
+- **Self-review trap breaker** — Concrete techniques to overcome shared mental model blind spots
+- **Structured verdicts** — BLOCK / CONCERNS / CLEAN with clear merge guidance
+
+## Usage
+
+```
+/adversarial-review              # Review staged/unstaged changes
+/adversarial-review --diff HEAD~3  # Review last 3 commits
+/adversarial-review --file src/auth.ts  # Review a specific file
+```
+
+## Examples
+
+### Example: Reviewing a PR Before Merge
+
+```
+/adversarial-review --diff main...HEAD
+```
+
+Produces a structured report with findings from all three personas, deduplicated and severity-ranked, ending with a BLOCK/CONCERNS/CLEAN verdict.
+
+## Problem This Solves
+
+When Claude reviews code it wrote (or code it just read), it shares the same mental model, assumptions, and blind spots as the author. This produces "Looks good to me" reviews on code that a fresh human reviewer would flag immediately. Users report this as one of the top frustrations with AI-assisted development.
+
+This skill forces a genuine perspective shift by requiring you to adopt adversarial personas — each with different priorities, different fears, and different definitions of "bad code."
+
+## Table of Contents
+
+1. [Quick Start](#quick-start)
+2. [Review Workflow](#review-workflow)
+3. [The Three Personas](#the-three-personas)
+4. [Severity Classification](#severity-classification)
+5. [Output Format](#output-format)
+6. [Anti-Patterns](#anti-patterns)
+7. [When to Use This](#when-to-use-this)
+
+## Quick Start
+
+```
+/adversarial-review              # Review staged/unstaged changes
+/adversarial-review --diff HEAD~3  # Review last 3 commits
+/adversarial-review --file src/auth.ts  # Review a specific file
+```
+
+## Review Workflow
+
+### Step 1: Gather the Changes
+
+Determine what to review based on invocation:
+
+- **No arguments:** Run `git diff` (unstaged) + `git diff --cached` (staged). If both empty, run `git diff HEAD~1` (last commit).
+- **`--diff <ref>`:** Run `git diff <ref>`.
+- **`--file <path>`:** Read the entire file. Focus review on the full file rather than just changes.
+
+If no changes are found, stop and report: "Nothing to review."
+
+### Step 2: Read the Full Context
+
+For every file in the diff:
+1. Read the **full file** (not just the changed lines) — bugs hide in how new code interacts with existing code.
+2. Identify the **purpose** of the change: bug fix, new feature, refactor, config change, test.
+3. Note any **project conventions** from CLAUDE.md, .editorconfig, linting configs, or existing patterns.
+
+### Step 3: Run All Three Personas
+
+Execute each persona sequentially. Each persona MUST produce at least one finding. If a persona finds nothing wrong, it has not looked hard enough — go back and look again.
+
+**IMPORTANT:** Do not soften findings. Do not hedge. Do not say "this might be fine but..." — either it's a problem or it isn't. Be direct.
+
+### Step 4: Deduplicate and Synthesize
+
+After all three personas have reported:
+1. Merge duplicate findings (same issue caught by multiple personas).
+2. Promote findings caught by 2+ personas to the next severity level.
+3. Produce the final structured output.
+
+## The Three Personas
+
+### Persona 1: The Saboteur
+
+**Mindset:** "I am trying to break this code in production."
+
+**Priorities:**
+- Input that was never validated
+- State that can become inconsistent
+- Concurrent access without synchronization
+- Error paths that swallow exceptions or return misleading results
+- Assumptions about data format, size, or availability that could be violated
+- Off-by-one errors, integer overflow, null/undefined dereferences
+- Resource leaks (file handles, connections, subscriptions, listeners)
+
+**Review Process:**
+1. For each function/method changed, ask: "What is the worst input I could send this?"
+2. For each external call, ask: "What if this fails, times out, or returns garbage?"
+3. For each state mutation, ask: "What if this runs twice? Concurrently? Never?"
+4. For each conditional, ask: "What if neither branch is correct?"
+
+**You MUST find at least one issue. If the code is genuinely bulletproof, note the most fragile assumption it relies on.**
+
+---
+
+### Persona 2: The New Hire
+
+**Mindset:** "I just joined this team. I need to understand and modify this code in 6 months with zero context from the original author."
+
+**Priorities:**
+- Names that don't communicate intent (what does `data` mean? what does `process()` do?)
+- Logic that requires reading 3+ other files to understand
+- Magic numbers, magic strings, unexplained constants
+- Functions doing more than one thing (the name says X but it also does Y and Z)
+- Missing type information that forces the reader to trace through call chains
+- Inconsistency with surrounding code style or project conventions
+- Tests that test implementation details instead of behavior
+- Comments that describe *what* (redundant) instead of *why* (useful)
+
+**Review Process:**
+1. Read each changed function as if you've never seen the codebase. Can you understand what it does from the name, parameters, and body alone?
+2. Trace one code path end-to-end. How many files do you need to open?
+3. Check: would a new contributor know where to add a similar feature?
+4. Look for "the author knew something the reader won't" — implicit knowledge baked into the code.
+
+**You MUST find at least one issue. If the code is crystal clear, note the most likely point of confusion for a newcomer.**
+
+---
+
+### Persona 3: The Security Auditor
+
+**Mindset:** "This code will be attacked. My job is to find the vulnerability before an attacker does."
+
+**OWASP-Informed Checklist:**
+
+| Category | What to Look For |
+|----------|-----------------|
+| **Injection** | SQL, NoSQL, OS command, LDAP — any place user input reaches a query or command without parameterization |
+| **Broken Auth** | Hardcoded credentials, missing auth checks on new endpoints, session tokens in URLs or logs |
+| **Data Exposure** | Sensitive data in error messages, logs, or API responses; missing encryption at rest or in transit |
+| **Insecure Defaults** | Debug mode left on, permissive CORS, wildcard permissions, default passwords |
+| **Missing Access Control** | IDOR (can user A access user B's data?), missing role checks, privilege escalation paths |
+| **Dependency Risk** | New dependencies with known CVEs, pinned to vulnerable versions, unnecessary transitive dependencies |
+| **Secrets** | API keys, tokens, passwords in code, config, or comments — even "temporary" ones |
+
+**Review Process:**
+1. Identify every trust boundary the code crosses (user input, API calls, database, file system, environment variables).
+2. For each boundary: is input validated? Is output sanitized? Is the principle of least privilege followed?
+3. Check: could an authenticated user escalate privileges through this change?
+4. Check: does this change expose any new attack surface?
+
+**You MUST find at least one issue. If the code has no security surface, note the closest thing to a security-relevant assumption.**
+
+## Severity Classification
+
+| Severity | Definition | Action Required |
+|----------|-----------|-----------------|
+| **CRITICAL** | Will cause data loss, security breach, or production outage. Must fix before merge. | Block merge. |
+| **WARNING** | Likely to cause bugs in edge cases, degrade performance, or confuse future maintainers. Should fix before merge. | Fix or explicitly accept risk with justification. |
+| **NOTE** | Style issue, minor improvement opportunity, or documentation gap. Nice to fix. | Author's discretion. |
+
+**Promotion rule:** A finding flagged by 2+ personas is promoted one level (NOTE becomes WARNING, WARNING becomes CRITICAL).
+
+## Output Format
+
+Structure your review as follows:
+
+```markdown
+## Adversarial Review: [brief description of what was reviewed]
+
+**Scope:** [files reviewed, lines changed, type of change]
+**Verdict:** BLOCK / CONCERNS / CLEAN
+
+### Critical Findings
+[If any — these block the merge]
+
+### Warnings
+[Should-fix items]
+
+### Notes
+[Nice-to-fix items]
+
+### Summary
+[2-3 sentences: what's the overall risk profile? What's the single most important thing to fix?]
+```
+
+**Verdict definitions:**
+- **BLOCK** — 1+ CRITICAL findings. Do not merge until resolved.
+- **CONCERNS** — No criticals but 2+ warnings. Merge at your own risk.
+- **CLEAN** — Only notes. Safe to merge.
+
+## Anti-Patterns
+
+### What This Skill is NOT
+
+| Anti-Pattern | Why It's Wrong |
+|-------------|---------------|
+| "LGTM, no issues found" | If you found nothing, you didn't look hard enough. Every change has at least one risk, assumption, or improvement opportunity. |
+| Cosmetic-only findings | Reporting only whitespace/formatting while missing a null dereference is worse than no review at all. Substance first, style second. |
+| Pulling punches | "This might possibly be a minor concern..." — No. Be direct. "This will throw a NullPointerException when `user` is undefined." |
+| Restating the diff | "This function was added to handle authentication" is not a finding. What's WRONG with how it handles authentication? |
+| Ignoring test gaps | New code without tests is a finding. Always. Tests are not optional. |
+| Reviewing only the changed lines | Bugs live in the interaction between new code and existing code. Read the full file. |
+
+### The Self-Review Trap
+
+You are likely reviewing code you just wrote or just read. Your brain (weights) formed the same mental model that produced this code. You will naturally think it looks correct because it matches your expectations.
+
+**To break this pattern:**
+1. Read the code **bottom-up** (start from the last function, work backward).
+2. For each function, state its contract **before** reading the body. Does the body match?
+3. Assume every variable could be null/undefined until proven otherwise.
+4. Assume every external call will fail.
+5. Ask: "If I deleted this change entirely, what would break?" — if the answer is "nothing," the change might be unnecessary.
+
+## When to Use This
+
+- **Before merging any PR** — especially self-authored PRs with no human reviewer
+- **After a long coding session** — fatigue produces blind spots; this skill compensates
+- **When Claude said "looks good"** — if you got an easy approval, run this for a second opinion
+- **On security-sensitive code** — auth, payments, data access, API endpoints
+- **When something "feels off"** — trust that instinct and run an adversarial review
+
+## Cross-References
+
+- Related: `engineering-team/senior-security` — deep security analysis
+- Related: `engineering-team/code-reviewer` — general code quality review
+- Complementary: `ra-qm-team/` — quality management workflows

package/bundled-skills/agent-device/SKILL.md

@@ -0,0 +1,20 @@
+---
+name: agent-device
+description: Automates Apple-platform apps (iOS, tvOS, macOS), Android devices, and Amazon Vega OS TV apps in Vega Virtual Devices. Use when navigating apps, taking snapshots/screenshots where supported, driving TV remotes, tapping, typing, scrolling, extracting UI info, collecting evidence, or planning agent-device CLI commands.
+---
+
+# agent-device
+
+For a normal app-driving task, start immediately. Do not probe first with `--help`, `--version`, `devices`, `appstate`, `snapshot`, or `screenshot`:
+
+```bash
+agent-device open <app> --foreground
+```
+
+That starts the session and returns the initial interactive snapshot with `@refs`.
+
+Loop: act with `press|click|fill|longpress <target> ... --settle`, `scroll <direction> --settle`, or `back --settle`; continue from the printed diff, verify the named expectation (`wait text "..."`, `is`, `get`, or `find`), then run `agent-device close`.
+
+Copy refs byte-for-byte: `@e12`, `@e12~s4` — keep the `@` and any `~sN`. Prefer current refs, then `id`/`label`/`role` selectors; coordinates are a last resort. If snapshot reports sparse/AX-unavailable, its refs and selectors are invalid: run `agent-device screenshot`, inspect the image, use coordinates, then retry `snapshot -i` after navigating. Otherwise run `snapshot -i` only when the diff lacks the next target.
+
+Error output includes corrective hints; follow them instead of re-planning. Only when the task is specialized (for example gestures, scripting, TV, macOS, remote, or debugging) or a command shape is unclear, run `agent-device help <topic>`. `agent-device --help` lists topics, but is not a startup step.