@allenpan2026/harshjudge 0.4.3 → 0.4.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (3) hide show
  1. package/.claude-plugin/plugin.json +1 -1
  2. package/package.json +1 -1
  3. package/skills/harshjudge/SKILL.md +24 -15
package/.claude-plugin/plugin.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "harshjudge",
3
- "version": "0.4.3",
3
+ "version": "0.4.4",
4
4
  "description": "AI-native E2E testing orchestration for Claude Code",
5
5
  "author": {
6
6
  "name": "Allen Pan"
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@allenpan2026/harshjudge",
3
- "version": "0.4.3",
3
+ "version": "0.4.4",
4
4
  "description": "AI-native E2E testing orchestration CLI for Claude Code.",
5
5
  "type": "module",
6
6
  "main": "./dist/index.js",
package/skills/harshjudge/SKILL.md CHANGED
@@ -108,18 +108,27 @@ Main Agent Step Agents (spawned per step)
108
108
  | `harshjudge discover search <pattern>` | Search file content |
109
109
  | `harshjudge dashboard open/close/status` | Manage dashboard server |
110
110
 
111
- ### Playwright MCP Tools
112
-
113
- | Tool | Purpose |
114
- |------|---------|
115
- | `browser_navigate` | Navigate to URL |
116
- | `browser_snapshot` | Get accessibility tree (use before click/type) |
117
- | `browser_click` | Click element using ref |
118
- | `browser_type` | Type into input using ref |
119
- | `browser_take_screenshot` | Capture screenshot for evidence |
120
- | `browser_console_messages` | Get console logs |
121
- | `browser_network_requests` | Get network activity |
122
- | `browser_wait_for` | Wait for text/condition |
111
+ ### Browser Automation (Auto-Detect)
112
+
113
+ Before running a scenario, detect which browser tool is available by checking for these tools in order:
114
+
115
+ 1. **Playwright MCP** — look for `browser_navigate` tool
116
+ 2. **browser-use** — look for `browser_use` tool or `browser-use` CLI
117
+ 3. **cmux-browser** — look for cmux browser surfaces
118
+
119
+ Use whichever is found first. The step agent needs these actions:
120
+
121
+ | Action | What to do |
122
+ |--------|-----------|
123
+ | Navigate | Go to a URL |
124
+ | Inspect | Get page state before interacting |
125
+ | Click | Click element by text/role/ref |
126
+ | Type | Enter text into input |
127
+ | Screenshot | Capture page as image file |
128
+ | Wait | Wait for text/element/timeout |
129
+ | Console | Read browser console output |
130
+
131
+ See [run-browser.md](references/run-browser.md) for tool-specific syntax.
123
132
 
124
133
  ## Step Agent Prompt Template
125
134
 
@@ -139,9 +148,9 @@ Auth: {from prd.md if needed}
139
148
  Status: {pass|fail|first step}
140
149
 
141
150
  ## Your Task
142
- 1. Execute the actions using Playwright MCP tools
143
- 2. Use browser_snapshot before clicking to get element refs
144
- 3. Capture before/after screenshots using browser_take_screenshot
151
+ 1. Execute the actions using the available browser tool
152
+ 2. Inspect the page before clicking or typing
153
+ 3. Capture before/after screenshots
145
154
  4. Record evidence: harshjudge evidence <runId> --step {stepNumber} --type screenshot --name before --data /path/to/screenshot.png
146
155
 
147
156
  Return ONLY a JSON object: