monomind 1.10.0 → 1.10.1

package/.claude/skills/agent-browser-testing/SKILL.md

@@ -1,312 +0,0 @@
----
-name: agent-browser-testing
-description: UI testing and task walkthrough using agent-browser — install, navigate, test golden paths, report issues, and help users accomplish tasks through any web UI
-version: 1.0.0
-triggers:
-  - /ui-test
-  - /browse
-  - /crawl
-  - ui test
-  - test the UI
-  - browser test
-  - test this system
-  - walk through
-  - check the UI
-  - QA this
-  - test frontend
-  - visual test
-  - crawl
-  - scrape
-  - browse the website
-  - go to the website
-  - open the website
-  - navigate to
-  - click on the website
-  - do this on the website
-  - automate the browser
-  - fill out the form
-  - log into
-  - sign in to
-  - submit the form
-tools:
-  - Bash
-requires:
-  - agent-browser >= 0.25.4
----
-
-# UI Testing with agent-browser
-
-Automated UI testing and guided task walkthroughs using `agent-browser`. Use this skill whenever a system's UI needs to be tested, explored, or used to accomplish a task.
-
-## Setup (Run Once)
-
-```bash
-# Install or upgrade agent-browser
-npm install -g agent-browser
-
-# Verify install
-agent-browser --version
-```
-
-## Core Testing Workflow
-
-Every UI test follows this loop:
-
-```
-OPEN → SNAPSHOT → ACT → SNAPSHOT → VERIFY → REPEAT
-```
-
-```bash
-# 1. Open the target UI
-agent-browser open <url>
-
-# 2. Get interactive elements (93% less context than full DOM)
-agent-browser snapshot -i
-
-# 3. Act using element refs from snapshot output
-agent-browser click @e1        # click by ref
-agent-browser fill @e2 "value" # fill input by ref
-agent-browser press Enter       # keyboard actions
-
-# 4. Re-snapshot to see result
-agent-browser snapshot -i
-
-# 5. Verify expected state
-agent-browser get text @e5      # read content
-agent-browser get url           # check URL changed
-agent-browser wait --text "Success"  # wait for expected text
-```
-
-## Test Phases
-
-### Phase 1 — Discovery
-```bash
-agent-browser open <url>
-agent-browser snapshot          # full tree to understand structure
-agent-browser get title
-agent-browser get url
-```
-
-### Phase 2 — Golden Path Testing
-Test the main happy-path flows a user would take:
-
-```bash
-# Example: Login flow
-agent-browser open https://app.example.com/login
-agent-browser snapshot -i
-# Identify: email input @e1, password @e2, submit @e3
-agent-browser fill @e1 "test@example.com"
-agent-browser fill @e2 "password123"
-agent-browser click @e3
-agent-browser wait --url "**/dashboard"
-agent-browser snapshot -i
-# Verify dashboard loaded
-```
-
-### Phase 3 — Edge Case Testing
-```bash
-# Empty form submission
-agent-browser click @e3              # submit with empty fields
-agent-browser wait --text "required" # expect validation error
-agent-browser snapshot -i
-
-# Invalid input
-agent-browser fill @e1 "not-an-email"
-agent-browser click @e3
-agent-browser snapshot -i
-
-# Boundary values
-agent-browser fill @e1 ""           # empty
-agent-browser fill @e1 "a".repeat(256)  # too long (via JS)
-```
-
-### Phase 4 — Navigation & Accessibility
-```bash
-# Tab through all focusable elements
-agent-browser press Tab
-agent-browser snapshot -i
-agent-browser press Tab
-# ... repeat checking focus state
-
-# Check all links work
-agent-browser find role link
-agent-browser snapshot -i
-
-# Check page at mobile width
-agent-browser resize 375 812
-agent-browser snapshot -i
-```
-
-### Phase 5 — Report Issues
-After testing, summarize:
-```
-✅ PASS: <what worked>
-❌ FAIL: <what broke> — steps to reproduce
-⚠️  WARN: <what looks odd but didn't break>
-```
-
-## Common Test Patterns
-
-### Login / Auth
-```bash
-agent-browser open <login-url>
-agent-browser snapshot -i
-agent-browser fill @e[email-input] "user@test.com"
-agent-browser fill @e[password-input] "TestPass123!"
-agent-browser click @e[submit-button]
-agent-browser wait --url "**/dashboard" --timeout 5000
-```
-
-### Form Submission
-```bash
-agent-browser open <form-url>
-agent-browser snapshot -i
-# Fill all required fields using refs from snapshot
-agent-browser fill @e1 "John Doe"
-agent-browser fill @e2 "john@test.com"
-agent-browser select @e3 "Option A"
-agent-browser check @e4              # checkbox
-agent-browser click @e5              # submit
-agent-browser wait --text "submitted"
-agent-browser screenshot test-result.png
-```
-
-### Multi-Step Wizard
-```bash
-# Step 1
-agent-browser open <wizard-url>
-agent-browser snapshot -i
-agent-browser fill @e1 "value"
-agent-browser click @e[next]
-agent-browser wait --text "Step 2"
-
-# Step 2
-agent-browser snapshot -i
-agent-browser select @e2 "choice"
-agent-browser click @e[next]
-
-# Final step — verify summary
-agent-browser snapshot -i
-agent-browser get text @e[summary]
-agent-browser click @e[confirm]
-agent-browser wait --text "Complete"
-```
-
-### CRUD Operations
-```bash
-# Create
-agent-browser click @e[add-button]
-agent-browser fill @e[name-field] "New Item"
-agent-browser click @e[save]
-agent-browser wait --text "New Item"  # appears in list
-
-# Read
-agent-browser get text @e[item-name]
-
-# Update
-agent-browser click @e[edit-button]
-agent-browser fill @e[name-field] "Updated Item"
-agent-browser click @e[save]
-
-# Delete
-agent-browser click @e[delete-button]
-agent-browser wait --text "Are you sure"
-agent-browser click @e[confirm-delete]
-agent-browser wait --not-text "Updated Item"
-```
-
-### API + UI Combination
-```bash
-# Trigger action via UI, verify via snapshot
-agent-browser click @e[send-button]
-agent-browser wait --text "Sent" --timeout 10000
-agent-browser screenshot after-send.png
-```
-
-## Selectors Reference
-
-Prefer element refs from snapshots — they're deterministic:
-```bash
-agent-browser snapshot -i
-# Output: button "Submit" [ref=e4]
-agent-browser click @e4  # use the ref
-```
-
-Fallback selectors:
-```bash
-agent-browser click "#submit-btn"           # CSS id
-agent-browser fill ".email-input" "test"    # CSS class
-agent-browser find role button click --name "Submit"   # ARIA role
-agent-browser find label "Email" fill "test"           # label text
-agent-browser find testid "submit-btn" click           # data-testid
-```
-
-## Task Walkthrough Mode
-
-When helping a user accomplish a task in a UI:
-
-1. **Ask for the URL** if not provided
-2. **Open and snapshot** to understand what's on screen
-3. **Describe what you see** — page title, main sections, available actions
-4. **Propose the steps** to accomplish the task
-5. **Execute step by step**, narrating each action
-6. **Confirm completion** — show what changed
-
-```bash
-# Example: "Help me create a new project in the app"
-agent-browser open https://app.example.com
-agent-browser snapshot -i
-# → "I can see: navbar with 'New Project' button at @e3, project list below"
-agent-browser click @e3
-agent-browser snapshot -i
-# → "Modal opened with: Name field @e8, Template selector @e9, Create button @e11"
-agent-browser fill @e8 "My New Project"
-agent-browser click @e11
-agent-browser wait --text "My New Project"
-# → "Project created successfully — it now appears in your project list"
-```
-
-## Screenshot & Evidence
-
-Always capture screenshots for test reports:
-```bash
-agent-browser screenshot before-action.png  # before
-agent-browser click @e[action]
-agent-browser screenshot after-action.png   # after
-
-# Full page screenshot
-agent-browser screenshot --full full-page.png
-```
-
-## Integration with Monomind
-
-### Store test patterns in memory
-```bash
-npx monomind memory store \
-  --namespace ui-testing \
-  --key "login-flow-<app-name>" \
-  --value "open→snapshot→fill @email →fill @password →click @submit →wait dashboard"
-```
-
-### Retrieve before re-testing
-```bash
-npx monomind memory search --query "login flow" --namespace ui-testing
-```
-
-### Report issues as tasks
-```bash
-npx monomind task create --title "UI Bug: form submits with empty email" \
-  --description "Steps: open /login, click submit without filling email — no validation shown"
-```
-
-## Activation Checklist
-
-When this skill is triggered:
-- [ ] Confirm `agent-browser` is installed (`agent-browser --version`)
-- [ ] Get the URL to test (ask user if not provided)
-- [ ] Open the URL and take initial snapshot
-- [ ] Identify the task or flow to test/accomplish
-- [ ] Execute the flow step by step
-- [ ] Report results (pass/fail/warnings)
-- [ ] Take screenshots of key states
-- [ ] Store successful patterns in memory for reuse