@weavetab/skills 2.5.1 → 2.6.0

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.
package/LICENSE CHANGED
@@ -1,6 +1,6 @@
1
1
  MIT License
2
2
 
3
- Copyright (c) 2026 fy2ne
3
+ Copyright (c) 2026 fy2ne <hello@fy2ne.me>
4
4
 
5
5
  Permission is hereby granted, free of charge, to any person obtaining a copy
6
6
  of this software and associated documentation files (the "Software"), to deal
package/README.md CHANGED
@@ -4,7 +4,7 @@
4
4
  [![GitHub stars](https://img.shields.io/github/stars/Weavetab/skills.svg?style=social)](https://github.com/Weavetab/skills)
5
5
  [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)
6
6
 
7
- Enterprise domain-driven agent intelligence for `@weavetab/mcp` (v2.5.1). Drop-in operational guides, failure recovery heuristics, and protocol rules for **Antigravity**, **Cursor**, **Cline**, **Windsurf**, and **OpenClaw**.
7
+ Enterprise domain-driven agent intelligence for `@weavetab/mcp` (v2.5.2). Drop-in operational guides, failure recovery heuristics, and protocol rules for **Antigravity**, **Cursor**, **Cline**, **Windsurf**, and **OpenClaw**.
8
8
 
9
9
  - **Website:** [weavetab.dev](https://weavetab.dev)
10
10
  - **Core MCP Server:** [`@weavetab/mcp`](https://www.npmjs.com/package/@weavetab/mcp)
@@ -55,7 +55,7 @@ Always select the narrowest tool that satisfies your immediate objective:
55
55
 
56
56
  ### Tier 1: Targeted Keyword Jump (`browser_find`)
57
57
  Use `browser_find` when you already know the label, text snippet, placeholder, or aria-label of the target element.
58
- - **Latency**: Sub-5ms CDP search.
58
+ - **Execution**: Leaf-optimized TreeWalker DOM query.
59
59
  - **Token Cost**: ~50–100 tokens (returns only matching candidate nodes).
60
60
 
61
61
  ```json
@@ -29,11 +29,11 @@ Submitting complex web forms one field at a time wastes agent turns and increase
29
29
  "fields": [
30
30
  {
31
31
  "ref": "w:12",
32
- "value": "Anas"
32
+ "value": "Alex"
33
33
  },
34
34
  {
35
35
  "ref": "w:14",
36
- "value": "Khezaz"
36
+ "value": "Rivera"
37
37
  },
38
38
  {
39
39
  "ref": "w:16",
@@ -144,8 +144,8 @@ Native operating system file picker dialogs block autonomous agent execution bec
144
144
  {
145
145
  "nodeId": 4821,
146
146
  "filePaths": [
147
- "C:/Users/me/Documents/contract.pdf",
148
- "C:/Users/me/Documents/financial_report.xlsx"
147
+ "./documents/contract.pdf",
148
+ "./documents/financial_report.xlsx"
149
149
  ]
150
150
  }
151
151
  ```
@@ -43,7 +43,7 @@ Pass `"app": "<application_name>"` to target the Electron application.
43
43
  {
44
44
  "app": "discord",
45
45
  "ref": "w:88",
46
- "text": "Deployment v2.5.1 completed successfully."
46
+ "text": "Deployment v2.5.2 completed successfully."
47
47
  }
48
48
  ```
49
49
 
@@ -44,13 +44,13 @@ Call `browser_detect` with empty parameters `{}`:
44
44
  {
45
45
  "name": "discord",
46
46
  "status": "running_no_debug",
47
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Discord\\app-1.0.9256\\Discord.exe"
47
+ "configuredPath": "%LOCALAPPDATA%\\Discord\\app-<version>\\Discord.exe"
48
48
  },
49
49
  {
50
50
  "name": "slack",
51
51
  "status": "connected",
52
52
  "port": 9222,
53
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\slack\\app-4.38.125\\slack.exe",
53
+ "configuredPath": "%LOCALAPPDATA%\\slack\\app-<version>\\slack.exe",
54
54
  "activeTabs": [
55
55
  {
56
56
  "id": "PAGE_1",
@@ -62,7 +62,7 @@ Call `browser_detect` with empty parameters `{}`:
62
62
  {
63
63
  "name": "vscode",
64
64
  "status": "offline",
65
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Programs\\Microsoft VS Code\\Code.exe"
65
+ "configuredPath": "%LOCALAPPDATA%\\Programs\\Microsoft VS Code\\Code.exe"
66
66
  }
67
67
  ],
68
68
  "toolCoverage": {
@@ -1,4 +1,4 @@
1
- # Weavetab Complete Enterprise Agent Suite (v2.5.1)
1
+ # Weavetab Complete Enterprise Agent Suite (v2.6.0)
2
2
 
3
3
 
4
4
 
@@ -8,7 +8,7 @@
8
8
 
9
9
  # Weavetab Domain Profile: BROWSER
10
10
 
11
- Generated for @weavetab/skills v2.5.1
11
+ Generated for @weavetab/skills v2.6.0
12
12
 
13
13
 
14
14
  ## dom-strategy
@@ -51,7 +51,7 @@ Always select the narrowest tool that satisfies your immediate objective:
51
51
 
52
52
  ### Tier 1: Targeted Keyword Jump (`browser_find`)
53
53
  Use `browser_find` when you already know the label, text snippet, placeholder, or aria-label of the target element.
54
- - **Latency**: Sub-5ms CDP search.
54
+ - **Execution**: Leaf-optimized TreeWalker DOM query.
55
55
  - **Token Cost**: ~50–100 tokens (returns only matching candidate nodes).
56
56
 
57
57
  ```json
@@ -134,11 +134,11 @@ Submitting complex web forms one field at a time wastes agent turns and increase
134
134
  "fields": [
135
135
  {
136
136
  "ref": "w:12",
137
- "value": "Anas"
137
+ "value": "Alex"
138
138
  },
139
139
  {
140
140
  "ref": "w:14",
141
- "value": "Khezaz"
141
+ "value": "Rivera"
142
142
  },
143
143
  {
144
144
  "ref": "w:16",
@@ -313,8 +313,8 @@ Native operating system file picker dialogs block autonomous agent execution bec
313
313
  {
314
314
  "nodeId": 4821,
315
315
  "filePaths": [
316
- "C:/Users/me/Documents/contract.pdf",
317
- "C:/Users/me/Documents/financial_report.xlsx"
316
+ "./documents/contract.pdf",
317
+ "./documents/financial_report.xlsx"
318
318
  ]
319
319
  }
320
320
  ```
@@ -540,7 +540,7 @@ The auto-heal routine re-establishes CDP protocol sessions, reconnects event lis
540
540
 
541
541
  # Weavetab Domain Profile: DESKTOP
542
542
 
543
- Generated for @weavetab/skills v2.5.1
543
+ Generated for @weavetab/skills v2.6.0
544
544
 
545
545
 
546
546
  ## desktop-targeting
@@ -570,7 +570,7 @@ Pass `"app": "<application_name>"` to target the Electron application.
570
570
  {
571
571
  "app": "discord",
572
572
  "ref": "w:88",
573
- "text": "Deployment v2.5.1 completed successfully."
573
+ "text": "Deployment v2.5.2 completed successfully."
574
574
  }
575
575
  ```
576
576
 
@@ -645,13 +645,13 @@ Call `browser_detect` with empty parameters `{}`:
645
645
  {
646
646
  "name": "discord",
647
647
  "status": "running_no_debug",
648
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Discord\\app-1.0.9256\\Discord.exe"
648
+ "configuredPath": "%LOCALAPPDATA%\\Discord\\app-<version>\\Discord.exe"
649
649
  },
650
650
  {
651
651
  "name": "slack",
652
652
  "status": "connected",
653
653
  "port": 9222,
654
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\slack\\app-4.38.125\\slack.exe",
654
+ "configuredPath": "%LOCALAPPDATA%\\slack\\app-<version>\\slack.exe",
655
655
  "activeTabs": [
656
656
  {
657
657
  "id": "PAGE_1",
@@ -663,7 +663,7 @@ Call `browser_detect` with empty parameters `{}`:
663
663
  {
664
664
  "name": "vscode",
665
665
  "status": "offline",
666
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Programs\\Microsoft VS Code\\Code.exe"
666
+ "configuredPath": "%LOCALAPPDATA%\\Programs\\Microsoft VS Code\\Code.exe"
667
667
  }
668
668
  ],
669
669
  "toolCoverage": {
@@ -722,7 +722,7 @@ Once launched with `--remote-debugging-port`, `browser_detect` immediately refle
722
722
 
723
723
  # Weavetab Domain Profile: MEMORY
724
724
 
725
- Generated for @weavetab/skills v2.5.1
725
+ Generated for @weavetab/skills v2.6.0
726
726
 
727
727
 
728
728
  ## selector-memory-system
@@ -1113,7 +1113,7 @@ Total agent turns for a known site: **3 tool calls** vs. potentially 12+ with fu
1113
1113
 
1114
1114
  # Weavetab Domain Profile: RESILIENCE
1115
1115
 
1116
- Generated for @weavetab/skills v2.5.1
1116
+ Generated for @weavetab/skills v2.6.0
1117
1117
 
1118
1118
 
1119
1119
  ## loop-breaker
@@ -1215,7 +1215,7 @@ If a downstream step errors out, times out, or hits an invalid state, roll back
1215
1215
  "name": "pre_checkout_authenticated"
1216
1216
  }
1217
1217
  ```
1218
- *Latency: ~15ms*. The page navigates back to the checkpoint URL with cookies and storage atomically injected.
1218
+ The page navigates back to the checkpoint URL with cookies and storage atomically injected.
1219
1219
 
1220
1220
  ### C. Listing Checkpoints (`list`)
1221
1221
  Inspect all active checkpoints for the current page target:
@@ -1318,7 +1318,7 @@ In modern Single Page Applications (built with React, Next.js, Vue, Nuxt, Svelte
1318
1318
 
1319
1319
  ## 1. The Multi-Tier Auto-Recovery Heuristic
1320
1320
 
1321
- When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs a sub-5ms heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
1321
+ When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs an in-memory heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
1322
1322
 
1323
1323
  ```
1324
1324
  [Action on w:14 fails / handle detached]
@@ -1363,7 +1363,7 @@ Either way:
1363
1363
 
1364
1364
  # Weavetab Domain Profile: SECURITY
1365
1365
 
1366
- Generated for @weavetab/skills v2.5.1
1366
+ Generated for @weavetab/skills v2.6.0
1367
1367
 
1368
1368
 
1369
1369
  ## rbac-and-policies
@@ -1462,7 +1462,7 @@ When `browser_type_secret` is invoked:
1462
1462
 
1463
1463
  # Weavetab Domain Profile: PLUGINS
1464
1464
 
1465
- Generated for @weavetab/skills v2.5.1
1465
+ Generated for @weavetab/skills v2.6.0
1466
1466
 
1467
1467
 
1468
1468
  ## hooks-and-overrides
@@ -1556,7 +1556,7 @@ Unverified plugins are loaded but their permissions are logged visibly at startu
1556
1556
 
1557
1557
  # Weavetab Domain Profile: NETWORK-AND-PERF
1558
1558
 
1559
- Generated for @weavetab/skills v2.5.1
1559
+ Generated for @weavetab/skills v2.6.0
1560
1560
 
1561
1561
 
1562
1562
  ## device-and-environment
@@ -1812,7 +1812,7 @@ Always synchronize with deterministic browser conditions:
1812
1812
 
1813
1813
  # Weavetab Domain Profile: DOCUMENTS
1814
1814
 
1815
- Generated for @weavetab/skills v2.5.1
1815
+ Generated for @weavetab/skills v2.6.0
1816
1816
 
1817
1817
 
1818
1818
  ## canvas-and-visual-capture
@@ -2002,7 +2002,7 @@ Word and PDF extraction return text in `content` (truncated by `max_chars`). PDF
2002
2002
 
2003
2003
  # Weavetab Domain Profile: DEVELOPER
2004
2004
 
2005
- Generated for @weavetab/skills v2.5.1
2005
+ Generated for @weavetab/skills v2.6.0
2006
2006
 
2007
2007
 
2008
2008
  ## github-forensics
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: BROWSER
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## dom-strategy
@@ -43,7 +43,7 @@ Always select the narrowest tool that satisfies your immediate objective:
43
43
 
44
44
  ### Tier 1: Targeted Keyword Jump (`browser_find`)
45
45
  Use `browser_find` when you already know the label, text snippet, placeholder, or aria-label of the target element.
46
- - **Latency**: Sub-5ms CDP search.
46
+ - **Execution**: Leaf-optimized TreeWalker DOM query.
47
47
  - **Token Cost**: ~50–100 tokens (returns only matching candidate nodes).
48
48
 
49
49
  ```json
@@ -126,11 +126,11 @@ Submitting complex web forms one field at a time wastes agent turns and increase
126
126
  "fields": [
127
127
  {
128
128
  "ref": "w:12",
129
- "value": "Anas"
129
+ "value": "Alex"
130
130
  },
131
131
  {
132
132
  "ref": "w:14",
133
- "value": "Khezaz"
133
+ "value": "Rivera"
134
134
  },
135
135
  {
136
136
  "ref": "w:16",
@@ -305,8 +305,8 @@ Native operating system file picker dialogs block autonomous agent execution bec
305
305
  {
306
306
  "nodeId": 4821,
307
307
  "filePaths": [
308
- "C:/Users/me/Documents/contract.pdf",
309
- "C:/Users/me/Documents/financial_report.xlsx"
308
+ "./documents/contract.pdf",
309
+ "./documents/financial_report.xlsx"
310
310
  ]
311
311
  }
312
312
  ```
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: DESKTOP
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## desktop-targeting
@@ -30,7 +30,7 @@ Pass `"app": "<application_name>"` to target the Electron application.
30
30
  {
31
31
  "app": "discord",
32
32
  "ref": "w:88",
33
- "text": "Deployment v2.5.1 completed successfully."
33
+ "text": "Deployment v2.5.2 completed successfully."
34
34
  }
35
35
  ```
36
36
 
@@ -105,13 +105,13 @@ Call `browser_detect` with empty parameters `{}`:
105
105
  {
106
106
  "name": "discord",
107
107
  "status": "running_no_debug",
108
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Discord\\app-1.0.9256\\Discord.exe"
108
+ "configuredPath": "%LOCALAPPDATA%\\Discord\\app-<version>\\Discord.exe"
109
109
  },
110
110
  {
111
111
  "name": "slack",
112
112
  "status": "connected",
113
113
  "port": 9222,
114
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\slack\\app-4.38.125\\slack.exe",
114
+ "configuredPath": "%LOCALAPPDATA%\\slack\\app-<version>\\slack.exe",
115
115
  "activeTabs": [
116
116
  {
117
117
  "id": "PAGE_1",
@@ -123,7 +123,7 @@ Call `browser_detect` with empty parameters `{}`:
123
123
  {
124
124
  "name": "vscode",
125
125
  "status": "offline",
126
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Programs\\Microsoft VS Code\\Code.exe"
126
+ "configuredPath": "%LOCALAPPDATA%\\Programs\\Microsoft VS Code\\Code.exe"
127
127
  }
128
128
  ],
129
129
  "toolCoverage": {
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: DEVELOPER
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## github-forensics
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: DOCUMENTS
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## canvas-and-visual-capture
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: MEMORY
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## selector-memory-system
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: NETWORK-AND-PERF
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## device-and-environment
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: PLUGINS
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## hooks-and-overrides
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: RESILIENCE
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## loop-breaker
@@ -102,7 +102,7 @@ If a downstream step errors out, times out, or hits an invalid state, roll back
102
102
  "name": "pre_checkout_authenticated"
103
103
  }
104
104
  ```
105
- *Latency: ~15ms*. The page navigates back to the checkpoint URL with cookies and storage atomically injected.
105
+ The page navigates back to the checkpoint URL with cookies and storage atomically injected.
106
106
 
107
107
  ### C. Listing Checkpoints (`list`)
108
108
  Inspect all active checkpoints for the current page target:
@@ -205,7 +205,7 @@ In modern Single Page Applications (built with React, Next.js, Vue, Nuxt, Svelte
205
205
 
206
206
  ## 1. The Multi-Tier Auto-Recovery Heuristic
207
207
 
208
- When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs a sub-5ms heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
208
+ When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs an in-memory heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
209
209
 
210
210
  ```
211
211
  [Action on w:14 fails / handle detached]
@@ -1,6 +1,6 @@
1
1
  # Weavetab Domain Profile: SECURITY
2
2
 
3
- Generated for @weavetab/skills v2.5.1
3
+ Generated for @weavetab/skills v2.6.0
4
4
 
5
5
 
6
6
  ## rbac-and-policies
@@ -55,7 +55,7 @@ Always select the narrowest tool that satisfies your immediate objective:
55
55
 
56
56
  ### Tier 1: Targeted Keyword Jump (`browser_find`)
57
57
  Use `browser_find` when you already know the label, text snippet, placeholder, or aria-label of the target element.
58
- - **Latency**: Sub-5ms CDP search.
58
+ - **Execution**: Leaf-optimized TreeWalker DOM query.
59
59
  - **Token Cost**: ~50–100 tokens (returns only matching candidate nodes).
60
60
 
61
61
  ```json
@@ -29,11 +29,11 @@ Submitting complex web forms one field at a time wastes agent turns and increase
29
29
  "fields": [
30
30
  {
31
31
  "ref": "w:12",
32
- "value": "Anas"
32
+ "value": "Alex"
33
33
  },
34
34
  {
35
35
  "ref": "w:14",
36
- "value": "Khezaz"
36
+ "value": "Rivera"
37
37
  },
38
38
  {
39
39
  "ref": "w:16",
@@ -144,8 +144,8 @@ Native operating system file picker dialogs block autonomous agent execution bec
144
144
  {
145
145
  "nodeId": 4821,
146
146
  "filePaths": [
147
- "C:/Users/me/Documents/contract.pdf",
148
- "C:/Users/me/Documents/financial_report.xlsx"
147
+ "./documents/contract.pdf",
148
+ "./documents/financial_report.xlsx"
149
149
  ]
150
150
  }
151
151
  ```
@@ -43,7 +43,7 @@ Pass `"app": "<application_name>"` to target the Electron application.
43
43
  {
44
44
  "app": "discord",
45
45
  "ref": "w:88",
46
- "text": "Deployment v2.5.1 completed successfully."
46
+ "text": "Deployment v2.5.2 completed successfully."
47
47
  }
48
48
  ```
49
49
 
@@ -44,13 +44,13 @@ Call `browser_detect` with empty parameters `{}`:
44
44
  {
45
45
  "name": "discord",
46
46
  "status": "running_no_debug",
47
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Discord\\app-1.0.9256\\Discord.exe"
47
+ "configuredPath": "%LOCALAPPDATA%\\Discord\\app-<version>\\Discord.exe"
48
48
  },
49
49
  {
50
50
  "name": "slack",
51
51
  "status": "connected",
52
52
  "port": 9222,
53
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\slack\\app-4.38.125\\slack.exe",
53
+ "configuredPath": "%LOCALAPPDATA%\\slack\\app-<version>\\slack.exe",
54
54
  "activeTabs": [
55
55
  {
56
56
  "id": "PAGE_1",
@@ -62,7 +62,7 @@ Call `browser_detect` with empty parameters `{}`:
62
62
  {
63
63
  "name": "vscode",
64
64
  "status": "offline",
65
- "configuredPath": "C:\\Users\\username\\AppData\\Local\\Programs\\Microsoft VS Code\\Code.exe"
65
+ "configuredPath": "%LOCALAPPDATA%\\Programs\\Microsoft VS Code\\Code.exe"
66
66
  }
67
67
  ],
68
68
  "toolCoverage": {
@@ -52,7 +52,7 @@ If a downstream step errors out, times out, or hits an invalid state, roll back
52
52
  "name": "pre_checkout_authenticated"
53
53
  }
54
54
  ```
55
- *Latency: ~15ms*. The page navigates back to the checkpoint URL with cookies and storage atomically injected.
55
+ The page navigates back to the checkpoint URL with cookies and storage atomically injected.
56
56
 
57
57
  ### C. Listing Checkpoints (`list`)
58
58
  Inspect all active checkpoints for the current page target:
@@ -22,7 +22,7 @@ In modern Single Page Applications (built with React, Next.js, Vue, Nuxt, Svelte
22
22
 
23
23
  ## 1. The Multi-Tier Auto-Recovery Heuristic
24
24
 
25
- When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs a sub-5ms heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
25
+ When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs an in-memory heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
26
26
 
27
27
  ```
28
28
  [Action on w:14 fails / handle detached]
@@ -1,5 +1,5 @@
1
1
  {
2
- "version": "2.5.1",
2
+ "version": "2.5.2",
3
3
  "scenarios": [
4
4
  {
5
5
  "id": "spa_stale_ref_recovery",
package/package.json CHANGED
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "name": "@weavetab/skills",
3
- "version": "2.5.1",
4
- "description": "Supercharge your AI Agents with Weavetab MCP Skills. Domain-driven agent intelligence for Antigravity, Cursor, Cline, Windsurf, and more.",
3
+ "version": "2.6.0",
4
+ "description": "Domain-driven operational skills and execution playbooks for Weavetab MCP.",
5
5
  "author": {
6
6
  "name": "fy2ne",
7
7
  "email": "hello@fy2ne.me",
@@ -51,7 +51,7 @@
51
51
  "agentic"
52
52
  ],
53
53
  "peerDependencies": {
54
- "@weavetab/mcp": "2.5.1"
54
+ "@weavetab/mcp": "2.6.0"
55
55
  },
56
56
  "peerDependenciesMeta": {
57
57
  "@weavetab/mcp": {
@@ -52,7 +52,7 @@ If a downstream step errors out, times out, or hits an invalid state, roll back
52
52
  "name": "pre_checkout_authenticated"
53
53
  }
54
54
  ```
55
- *Latency: ~15ms*. The page navigates back to the checkpoint URL with cookies and storage atomically injected.
55
+ The page navigates back to the checkpoint URL with cookies and storage atomically injected.
56
56
 
57
57
  ### C. Listing Checkpoints (`list`)
58
58
  Inspect all active checkpoints for the current page target:
@@ -22,7 +22,7 @@ In modern Single Page Applications (built with React, Next.js, Vue, Nuxt, Svelte
22
22
 
23
23
  ## 1. The Multi-Tier Auto-Recovery Heuristic
24
24
 
25
- When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs a sub-5ms heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
25
+ When a tool receives a detached or invalidated `ref`, Weavetab attempts to re-resolve it instead of failing immediately. The recovery runs an in-memory heuristic over these strategies (in order) and re-stamps `data-wt-id="NN"` onto the live node:
26
26
 
27
27
  ```
28
28
  [Action on w:14 fails / handle detached]
package/skills.json CHANGED
@@ -1,5 +1,5 @@
1
1
  {
2
- "version": "2.5.1",
2
+ "version": "2.6.0",
3
3
  "domains": [
4
4
  "browser",
5
5
  "desktop",