@monoes/monomindcli 1.10.31 → 1.10.33

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

@@ -1,8 +1,7 @@
 ---
 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. Delegates to monomind:browse for full feature coverage.
-version: 2.0.0
-delegates_to: monomind:browse
+description: UI testing and task walkthrough using the native monomind browse command — navigate, test golden paths, report issues, and help users accomplish tasks through any web UI.
+version: 3.0.0
 triggers:
   - /ui-test
   - /browse
@@ -31,23 +30,11 @@ triggers:
   - submit the form
 tools:
   - Bash
-requires:
-  - agent-browser >= 0.25.4
 ---
 
-# UI Testing with agent-browser
+# UI Testing with monomind browse
 
-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
-```
+Automated UI testing and guided task walkthroughs using the native `monomind browse` command (TypeScript CDP client, no external binary required).
 
 ## Core Testing Workflow
 
@@ -59,187 +46,173 @@ OPEN → SNAPSHOT → ACT → SNAPSHOT → VERIFY → REPEAT
 
 ```bash
 # 1. Open the target UI
-agent-browser open <url>
+npx monomind browse open <url>
 
 # 2. Get interactive elements (93% less context than full DOM)
-agent-browser snapshot -i
+npx monomind browse 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
+npx monomind browse click @e1        # click by ref
+npx monomind browse fill @e2 "value" # fill input by ref
+npx monomind browse press Enter       # keyboard actions
 
 # 4. Re-snapshot to see result
-agent-browser snapshot -i
+npx monomind browse 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
+npx monomind browse get text @e5          # read element text
+npx monomind browse get url               # check URL changed
+npx monomind browse wait --text "Success" # wait for expected text
+npx monomind browse errors                # check for JS errors
 ```
 
 ## 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
+npx monomind browse open <url>
+npx monomind browse snapshot       # full tree to understand structure
+npx monomind browse get title
+npx monomind browse 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
+npx monomind browse open https://app.example.com/login
+npx monomind browse 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
+npx monomind browse fill @e1 "test@example.com"
+npx monomind browse fill @e2 "password123"
+npx monomind browse click @e3
+npx monomind browse wait --url "**/dashboard"
+npx monomind browse 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
+npx monomind browse click @e3              # submit with empty fields
+npx monomind browse wait --text "required" # expect validation error
+npx monomind browse 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)
+npx monomind browse fill @e1 "not-an-email"
+npx monomind browse click @e3
+npx monomind browse snapshot -i
+
+# Check element states
+npx monomind browse is visible @e5
+npx monomind browse is enabled @e3
+npx monomind browse is checked @e7
 ```
 
 ### 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
+npx monomind browse press Tab
+npx monomind browse snapshot -i
 
 # Check all links work
-agent-browser find role link
-agent-browser snapshot -i
+npx monomind browse find role link
 
 # Check page at mobile width
-agent-browser resize 375 812
-agent-browser snapshot -i
+npx monomind browse set device "iPhone 14"
+npx monomind browse snapshot -i
+
+# Check dark mode
+npx monomind browse set media dark
+npx monomind browse screenshot dark-mode.png
 ```
 
 ### Phase 5 — Report Issues
 After testing, summarize:
 ```
-✅ PASS: <what worked>
-❌ FAIL: <what broke> — steps to reproduce
-⚠️  WARN: <what looks odd but didn't break>
+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
+npx monomind browse open <login-url>
+npx monomind browse snapshot -i
+npx monomind browse fill @e1 "user@test.com"
+npx monomind browse fill @e2 "TestPass123!"
+npx monomind browse click @e3
+npx monomind browse wait --url "**/dashboard" --timeout 5000
+npx monomind browse errors
 ```
 
 ### 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
+npx monomind browse open <form-url>
+npx monomind browse snapshot -i
+npx monomind browse fill @e1 "John Doe"
+npx monomind browse fill @e2 "john@test.com"
+npx monomind browse select @e3 "Option A"
+npx monomind browse check @e4
+npx monomind browse click @e5
+npx monomind browse wait --text "submitted"
+npx monomind browse 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"
+npx monomind browse open <wizard-url>
+npx monomind browse snapshot -i
+npx monomind browse fill @e1 "value"
+npx monomind browse click @e2
+npx monomind browse wait --text "Step 2"
+
+npx monomind browse snapshot -i
+npx monomind browse select @e3 "choice"
+npx monomind browse click @e4
+
+npx monomind browse snapshot -i
+npx monomind browse get text @e5
+npx monomind browse click @e6
+npx monomind browse 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]
+npx monomind browse click @e1
+npx monomind browse fill @e2 "New Item"
+npx monomind browse click @e3
+npx monomind browse wait --text "New Item"
 
 # Update
-agent-browser click @e[edit-button]
-agent-browser fill @e[name-field] "Updated Item"
-agent-browser click @e[save]
+npx monomind browse find text "Edit" click
+npx monomind browse fill @e2 "Updated Item"
+npx monomind browse click @e3
 
 # 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
+npx monomind browse find text "Delete" click
+npx monomind browse wait --text "Are you sure"
+npx monomind browse click @e4
+npx monomind browse wait --load networkidle
 ```
 
 ## 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
+npx monomind browse snapshot -i
+# Output: button "Submit" [@e4]
+npx monomind browse click @e4
 ```
 
-Fallback selectors:
+Semantic locators (no CSS needed):
 ```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
+npx monomind browse find role button click --name "Submit"
+npx monomind browse find label "Email" fill "test@example.com"
+npx monomind browse find text "Delete" click
+npx monomind browse find testid "submit-btn" click
 ```
 
 ## Task Walkthrough Mode
@@ -254,60 +227,85 @@ When helping a user accomplish a task in a UI:
 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
+npx monomind browse open https://app.example.com
+npx monomind browse snapshot -i
 # → "I can see: navbar with 'New Project' button at @e3, project list below"
-agent-browser click @e3
-agent-browser snapshot -i
+npx monomind browse click @e3
+npx monomind browse 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"
+npx monomind browse fill @e8 "My New Project"
+npx monomind browse click @e11
+npx monomind browse 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
+npx monomind browse screenshot before-action.png
+npx monomind browse click @e1
+npx monomind browse screenshot after-action.png
 
 # Full page screenshot
-agent-browser screenshot --full full-page.png
+npx monomind browse screenshot --full full-page.png
 ```
 
-## Integration with Monomind
+## Advanced Capabilities
 
-### Store test patterns in memory
 ```bash
+# Console & error monitoring
+npx monomind browse console
+npx monomind browse errors
+
+# Network interception
+npx monomind browse network route --pattern "https://api.*" --abort
+
+# Device emulation
+npx monomind browse set device "iPhone 14"
+npx monomind browse set device "Galaxy S21"
+
+# Session persistence
+npx monomind browse state save my-session
+npx monomind browse open <url> --session my-session
+
+# Drag and drop
+npx monomind browse drag @e1 @e2
+
+# File upload
+npx monomind browse upload @e1 ./test-file.pdf
+
+# Cookie management
+npx monomind browse cookies list
+npx monomind browse cookies set --name token --value abc123
+
+# PDF export
+npx monomind browse pdf ./output.pdf
+
+# Batch execution
+npx monomind browse batch "open https://app.com" "snapshot -i" "click @e3"
+```
+
+## Integration with Monomind Memory
+
+```bash
+# Store successful test patterns
 npx monomind memory store \
   --namespace ui-testing \
   --key "login-flow-<app-name>" \
-  --value "open→snapshot→fill @email →fill @password →click @submit →wait dashboard"
-```
+  --value "open→snapshot -i→fill @e1 email→fill @e2 pass→click @e3→wait dashboard"
 
-### Retrieve before re-testing
-```bash
+# Retrieve before re-testing
 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
+- [ ] Run `npx monomind browse open <url>` then `snapshot -i`
 - [ ] Identify the task or flow to test/accomplish
 - [ ] Execute the flow step by step
+- [ ] Run `npx monomind browse errors` to check for JS failures
 - [ ] Report results (pass/fail/warnings)
 - [ ] Take screenshots of key states
 - [ ] Store successful patterns in memory for reuse

package/.claude/skills/monomind/browse-agentcore.md

@@ -11,10 +11,9 @@ triggers:
 tools:
   - Bash
 requires:
-  - agent-browser >= 0.25.4
+  - monomind >= 1.0.0
 ---
 
-<!-- Pattern adapted from agent-browser — rebranded for monomind -->
 
 # AWS Bedrock AgentCore Browser (monomind:browse-agentcore)
 
@@ -34,13 +33,13 @@ No additional setup needed if AWS credentials are already configured.
 
 ```bash
 # Open a page on an AgentCore cloud browser
-agent-browser -p agentcore open https://example.com
+npx monomind browse -p agentcore open https://example.com
 
 # Everything else is identical to local Chrome
-agent-browser snapshot -i
-agent-browser click @e1
-agent-browser screenshot page.png
-agent-browser close
+npx monomind browse snapshot -i
+npx monomind browse click @e1
+npx monomind browse screenshot page.png
+npx monomind browse close
 ```
 
 ## Environment Variables
@@ -59,10 +58,10 @@ agent-browser close
 export AGENT_BROWSER_PROVIDER=agentcore
 export AGENTCORE_REGION=us-east-2
 
-agent-browser open https://example.com
-agent-browser snapshot -i
-agent-browser click @e1
-agent-browser close
+npx monomind browse open https://example.com
+npx monomind browse snapshot -i
+npx monomind browse click @e1
+npx monomind browse close
 ```
 
 ## Persistent Profiles
@@ -71,15 +70,15 @@ Reuse login state across sessions:
 
 ```bash
 # First run — log in
-AGENTCORE_PROFILE_ID=my-app agent-browser -p agentcore open https://app.example.com/login
-agent-browser snapshot -i
-agent-browser fill @e1 "user@example.com"
-agent-browser fill @e2 "password"
-agent-browser click @e3
-agent-browser close
+AGENTCORE_PROFILE_ID=my-app monomind browse -p agentcore open https://app.example.com/login
+npx monomind browse snapshot -i
+npx monomind browse fill @e1 "user@example.com"
+npx monomind browse fill @e2 "password"
+npx monomind browse click @e3
+npx monomind browse close
 
 # Future runs — already authenticated
-AGENTCORE_PROFILE_ID=my-app agent-browser -p agentcore open https://app.example.com/dashboard
+AGENTCORE_PROFILE_ID=my-app monomind browse -p agentcore open https://app.example.com/dashboard
 ```
 
 ## Credential Patterns
@@ -88,14 +87,14 @@ AGENTCORE_PROFILE_ID=my-app agent-browser -p agentcore open https://app.example.
 # Explicit (CI/CD)
 export AWS_ACCESS_KEY_ID=AKIA...
 export AWS_SECRET_ACCESS_KEY=...
-agent-browser -p agentcore open https://example.com
+npx monomind browse -p agentcore open https://example.com
 
 # SSO
 aws sso login --profile my-profile
-AWS_PROFILE=my-profile agent-browser -p agentcore open https://example.com
+AWS_PROFILE=my-profile monomind browse -p agentcore open https://example.com
 
 # Default credential chain (IAM role, etc.)
-agent-browser -p agentcore open https://example.com
+npx monomind browse -p agentcore open https://example.com
 ```
 
 ## Live View