davidgasquez.com
🔧 Where my dotfiles lives in harmony and peace, most of the time
📝 Update agent settings and skill docs
+2
-2
agents/pi/settings.json
+2
-2
agents/pi/settings.json
+113
-5
agents/skills/agent-browser/SKILL.md
+113
-5
agents/skills/agent-browser/SKILL.md
···
35
35
36
36
### Snapshot (page analysis)
37
37
```bash
38
-
agent-browser snapshot # Full accessibility tree
39
-
agent-browser snapshot -i # Interactive elements only (recommended)
40
-
agent-browser snapshot -c # Compact output
41
-
agent-browser snapshot -d 3 # Limit depth to 3
38
+
agent-browser snapshot # Full accessibility tree
39
+
agent-browser snapshot -i # Interactive elements only (recommended)
40
+
agent-browser snapshot -c # Compact output
41
+
agent-browser snapshot -d 3 # Limit depth to 3
42
+
agent-browser snapshot -s "#main" # Scope to CSS selector
42
43
```
43
44
44
45
### Interactions (use @refs from snapshot)
45
46
```bash
46
47
agent-browser click @e1 # Click
47
48
agent-browser dblclick @e1 # Double-click
49
+
agent-browser focus @e1 # Focus element
48
50
agent-browser fill @e2 "text" # Clear and type
49
51
agent-browser type @e2 "text" # Type without clearing
50
52
agent-browser press Enter # Press key
51
53
agent-browser press Control+a # Key combination
54
+
agent-browser keydown Shift # Hold key down
55
+
agent-browser keyup Shift # Release key
52
56
agent-browser hover @e1 # Hover
53
57
agent-browser check @e1 # Check checkbox
54
58
agent-browser uncheck @e1 # Uncheck checkbox
55
59
agent-browser select @e1 "value" # Select dropdown
56
60
agent-browser scroll down 500 # Scroll page
57
61
agent-browser scrollintoview @e1 # Scroll element into view
62
+
agent-browser drag @e1 @e2 # Drag and drop
63
+
agent-browser upload @e1 file.pdf # Upload files
58
64
```
59
65
60
66
### Get information
61
67
```bash
62
68
agent-browser get text @e1 # Get element text
69
+
agent-browser get html @e1 # Get innerHTML
63
70
agent-browser get value @e1 # Get input value
71
+
agent-browser get attr @e1 href # Get attribute
64
72
agent-browser get title # Get page title
65
73
agent-browser get url # Get current URL
74
+
agent-browser get count ".item" # Count matching elements
75
+
agent-browser get box @e1 # Get bounding box
66
76
```
67
77
68
-
### Screenshots
78
+
### Check state
79
+
```bash
80
+
agent-browser is visible @e1 # Check if visible
81
+
agent-browser is enabled @e1 # Check if enabled
82
+
agent-browser is checked @e1 # Check if checked
83
+
```
84
+
85
+
### Screenshots & PDF
69
86
```bash
70
87
agent-browser screenshot # Screenshot to stdout
71
88
agent-browser screenshot path.png # Save to file
72
89
agent-browser screenshot --full # Full page
90
+
agent-browser pdf output.pdf # Save as PDF
73
91
```
74
92
93
+
### Video recording
94
+
```bash
95
+
agent-browser record start ./demo.webm # Start recording (uses current URL + state)
96
+
agent-browser click @e1 # Perform actions
97
+
agent-browser record stop # Stop and save video
98
+
agent-browser record restart ./take2.webm # Stop current + start new recording
99
+
```
100
+
Recording creates a fresh context but preserves cookies/storage from your session. If no URL is provided, it automatically returns to your current page. For smooth demos, explore first, then start recording.
101
+
75
102
### Wait
76
103
```bash
77
104
agent-browser wait @e1 # Wait for element
78
105
agent-browser wait 2000 # Wait milliseconds
79
106
agent-browser wait --text "Success" # Wait for text
107
+
agent-browser wait --url "**/dashboard" # Wait for URL pattern
80
108
agent-browser wait --load networkidle # Wait for network idle
109
+
agent-browser wait --fn "window.ready" # Wait for JS condition
110
+
```
111
+
112
+
### Mouse control
113
+
```bash
114
+
agent-browser mouse move 100 200 # Move mouse
115
+
agent-browser mouse down left # Press button
116
+
agent-browser mouse up left # Release button
117
+
agent-browser mouse wheel 100 # Scroll wheel
81
118
```
82
119
83
120
### Semantic locators (alternative to refs)
···
85
122
agent-browser find role button click --name "Submit"
86
123
agent-browser find text "Sign In" click
87
124
agent-browser find label "Email" fill "user@test.com"
125
+
agent-browser find first ".item" click
126
+
agent-browser find nth 2 "a" text
127
+
```
128
+
129
+
### Browser settings
130
+
```bash
131
+
agent-browser set viewport 1920 1080 # Set viewport size
132
+
agent-browser set device "iPhone 14" # Emulate device
133
+
agent-browser set geo 37.7749 -122.4194 # Set geolocation
134
+
agent-browser set offline on # Toggle offline mode
135
+
agent-browser set headers '{"X-Key":"v"}' # Extra HTTP headers
136
+
agent-browser set credentials user pass # HTTP basic auth
137
+
agent-browser set media dark # Emulate color scheme
138
+
```
139
+
140
+
### Cookies & Storage
141
+
```bash
142
+
agent-browser cookies # Get all cookies
143
+
agent-browser cookies set name value # Set cookie
144
+
agent-browser cookies clear # Clear cookies
145
+
agent-browser storage local # Get all localStorage
146
+
agent-browser storage local key # Get specific key
147
+
agent-browser storage local set k v # Set value
148
+
agent-browser storage local clear # Clear all
149
+
```
150
+
151
+
### Network
152
+
```bash
153
+
agent-browser network route <url> # Intercept requests
154
+
agent-browser network route <url> --abort # Block requests
155
+
agent-browser network route <url> --body '{}' # Mock response
156
+
agent-browser network unroute [url] # Remove routes
157
+
agent-browser network requests # View tracked requests
158
+
agent-browser network requests --filter api # Filter requests
159
+
```
160
+
161
+
### Tabs & Windows
162
+
```bash
163
+
agent-browser tab # List tabs
164
+
agent-browser tab new [url] # New tab
165
+
agent-browser tab 2 # Switch to tab
166
+
agent-browser tab close # Close tab
167
+
agent-browser window new # New window
168
+
```
169
+
170
+
### Frames
171
+
```bash
172
+
agent-browser frame "#iframe" # Switch to iframe
173
+
agent-browser frame main # Back to main frame
174
+
```
175
+
176
+
### Dialogs
177
+
```bash
178
+
agent-browser dialog accept [text] # Accept dialog
179
+
agent-browser dialog dismiss # Dismiss dialog
180
+
```
181
+
182
+
### JavaScript
183
+
```bash
184
+
agent-browser eval "document.title" # Run JavaScript
88
185
```
89
186
90
187
## Example: Form submission
···
137
234
## Debugging
138
235
139
236
```bash
237
+
agent-browser open example.com --headed # Show browser window
238
+
agent-browser console # View console messages
239
+
agent-browser errors # View page errors
240
+
agent-browser record start ./debug.webm # Record from current page
241
+
agent-browser record stop # Save recording
140
242
agent-browser open example.com --headed # Show browser window
243
+
agent-browser --cdp 9222 snapshot # Connect via CDP
141
244
agent-browser console # View console messages
245
+
agent-browser console --clear # Clear console
142
246
agent-browser errors # View page errors
247
+
agent-browser errors --clear # Clear errors
248
+
agent-browser highlight @e1 # Highlight element
249
+
agent-browser trace start # Start recording trace
250
+
agent-browser trace stop trace.zip # Stop and save trace
143
251
```
+2
-2
agents/skills/ask-questions-if-underspecified/SKILL.md
+2
-2
agents/skills/ask-questions-if-underspecified/SKILL.md
···
50
50
51
51
Once you have answers, restate the requirements in 1-3 sentences (including key constraints and what success looks like), then start work.
52
52
53
-
## Question templates
53
+
## Question Templates
54
54
55
55
- "Before I start, I need: (1) ..., (2) ..., (3) .... If you don't care about (2), I will assume ...."
56
56
- "Which of these should it be? A) ... B) ... C) ... (pick one)"
···
71
71
Reply with: defaults (or 1a 2a)
72
72
```
73
73
74
-
## Anti-patterns
74
+
## Anti-Patterns
75
75
76
76
- Don't ask questions you can answer with a quick, low-risk discovery read (e.g., configs, existing patterns, docs).
77
77
- Don't ask open-ended questions if a tight multiple-choice or yes/no would eliminate ambiguity faster.