upgrade dependencies and improve imports

Author: DenizOkcu

docs/tool-whitelist.md

@@ -15,6 +15,7 @@ Tools are capabilities that allow AI models to interact with your vault and the
 - **Web Search**: Search the web via Brave Search API
 
 **Privacy Guarantee**: All tool calls require your explicit approval at three layers:
+
 1. Approve the tool execution request
 2. Review the results before sharing with AI
 3. Select exactly what data to share
@@ -25,12 +26,12 @@ See [v3.0.0 Release Notes](https://github.com/obsidian-chatgpt-md/obsidian-chatg
 
 The default whitelist includes **161 patterns** covering models from live API testing:
 
-| Provider | Patterns | Models Tested | Tool Support |
-|----------|----------|---------------|--------------|
-| OpenAI | 36 | 119 | 56 models |
-| Anthropic | 9 | 9 | 9 models (100%) |
-| Gemini | 7 | 29 | 7 models |
-| OpenRouter | 109 | 346 | 122 models |
+| Provider   | Patterns | Models Tested | Tool Support    |
+| ---------- | -------- | ------------- | --------------- |
+| OpenAI     | 36       | 119           | 56 models       |
+| Anthropic  | 9        | 9             | 9 models (100%) |
+| Gemini     | 7        | 29            | 7 models        |
+| OpenRouter | 109      | 346           | 122 models      |
 
 **Test Date**: 2026-02-01
 **Total Tested**: 503 models
@@ -47,14 +48,17 @@ The whitelist uses flexible pattern matching to handle model versioning:
 ```
 gpt-4o
 ```
+
 Matches only `gpt-4o`
 
 ### Date Suffix Auto-Match
 
 ```
 gpt-4o
 ```
+
 Matches:
+
 - `gpt-4o`
 - `gpt-4o-2025-04-16`
 - `gpt-4o-20251101`
@@ -66,7 +70,9 @@ Date suffixes are automatically matched, so you don't need to list every version
 ```
 gpt-4*
 ```
+
 Matches anything starting with `gpt-4`:
+
 - `gpt-4`
 - `gpt-4-turbo`
 - `gpt-4o`
@@ -78,6 +84,7 @@ Matches anything starting with `gpt-4`:
 # This is a comment
 gpt-4o
 ```
+
 Lines starting with `#` are ignored.
 
 ## Configuration
@@ -185,6 +192,7 @@ toolEnabledModels: |
 ### OpenRouter
 
 **Tool Support**: 109 confirmed patterns including:
+
 - DeepSeek (V3, R1 series)
 - Qwen (2.5, 3 series)
 - Mistral (Large, Ministral)
@@ -215,10 +223,7 @@ Use the **WhitelistValidator** to see which of your configured models support to
 ```typescript
 import { validateWhitelist } from "src/Services/WhitelistValidator";
 
-const result = validateWhitelist(
-  ["gpt-4o", "claude-3-haiku", "unknown-model"],
-  whitelist
-);
+const result = validateWhitelist(["gpt-4o", "claude-3-haiku", "unknown-model"], whitelist);
 
 console.log(result.matchRate); // Percentage supporting tools
 console.log(result.unmatchedModelDetails); // Models not on whitelist
@@ -231,19 +236,22 @@ console.log(result.unmatchedModelDetails); // Models not on whitelist
 **Cause**: Model returned an error when tools were in the API request.
 
 **Solutions**:
+
 1. Check if model is actually on whitelist
 2. Try a confirmed model from default whitelist
 3. Enable Debug Mode to see API response
 
 #### Tools not offered when expected
 
 **Cause**: Multiple possible reasons:
+
 1. Tool calling disabled in settings
 2. Model not on whitelist
 3. Model doesn't understand it can use tools
 4. No relevant tools for the query
 
 **Solutions**:
+
 1. Check `enableToolCalling` is on
 2. Verify model is in whitelist
 3. Try explicit query: "Search my notes for..."
@@ -254,6 +262,7 @@ console.log(result.unmatchedModelDetails); // Models not on whitelist
 **Cause**: Whitelist needs full `provider/model` format.
 
 **Solution**:
+
 ```yaml
 # Correct
 toolEnabledModels: |
@@ -383,6 +392,7 @@ Note: The `!` exclusion syntax is not currently supported. You must explicitly l
 ### Q: Why isn't my model on the whitelist?
 
 **A**: The model either:
+
 1. Hasn't been tested yet (new models added regularly)
 2. Failed tool support testing (errors or inconsistent behavior)
 3. Is too old and doesn't support modern tool calling
@@ -396,6 +406,7 @@ Test it yourself and add it if it works!
 ### Q: How often is the whitelist updated?
 
 **A**: The whitelist is updated when:
+
 - Major providers release new models
 - Community testing confirms additional models
 - Bug reports indicate false positives/negatives
@@ -405,6 +416,7 @@ Check the test date in the whitelist header for last update.
 ### Q: What if a model stops supporting tools?
 
 **A**: This is rare but possible. If you encounter this:
+
 1. Enable Debug Mode to confirm the error
 2. Remove the model from your whitelist
 3. Report it as an issue

package.json

@@ -21,25 +21,25 @@
   "keywords": [],
   "author": "",
   "dependencies": {
-    "@ai-sdk/anthropic": "^3.0.1",
-    "@ai-sdk/google": "^3.0.1",
-    "@ai-sdk/openai": "^3.0.1",
-    "@ai-sdk/openai-compatible": "^2.0.1",
-    "@openrouter/ai-sdk-provider": "^1.5.4",
-    "ai": "6.0.3",
-    "zod": "4.2.1"
+    "@ai-sdk/anthropic": "^3.0.35",
+    "@ai-sdk/google": "^3.0.20",
+    "@ai-sdk/openai": "^3.0.25",
+    "@ai-sdk/openai-compatible": "^2.0.26",
+    "@openrouter/ai-sdk-provider": "^2.1.1",
+    "ai": "6.0.67",
+    "zod": "4.3.6"
   },
   "devDependencies": {
-    "@codemirror/state": "^6.5.3",
-    "@codemirror/view": "^6.39.7",
-    "@types/node": "^25.0.3",
-    "@typescript-eslint/eslint-plugin": "8.50.1",
-    "@typescript-eslint/parser": "8.50.1",
+    "@codemirror/state": "^6.5.4",
+    "@codemirror/view": "^6.39.12",
+    "@types/node": "^25.2.0",
+    "@typescript-eslint/eslint-plugin": "8.54.0",
+    "@typescript-eslint/parser": "8.54.0",
     "builtin-modules": "5.0.0",
     "esbuild": "0.27.2",
     "eslint": "^9.39.2",
     "obsidian": "latest",
-    "prettier": "^3.7.4",
+    "prettier": "^3.8.1",
     "tslib": "2.8.1",
     "typescript": "5.9.3"
   }

scripts/tool-whitelist/QUICK-START.md

@@ -65,10 +65,12 @@ node scripts/tool-whitelist/generate-whitelist.mjs
 ## Options Cheat Sheet
 
 ### test-models-tools.mjs
+
 - `--limit N` - Test only N models per provider
 - `--provider name` - Test only: openai|anthropic|gemini|openrouter
 
 ### generate-whitelist.mjs
+
 - `--min-success-rate N` - Minimum rate 0-1 (default: 1.0)
 - `--output path` - Custom output path
 

scripts/tool-whitelist/README-WHITELIST-MAINTENANCE.md

@@ -29,11 +29,13 @@ node scripts/tool-whitelist/fetch-available-models.mjs
 **Output**: `scripts/tool-whitelist/available-models.json`
 
 **What it does**:
+
 - Queries each provider's API for available models
 - Saves model metadata (id, name, created date, etc.)
 - No tool testing - just discovery
 
 **Example output**:
+
 ```json
 {
   "fetchedAt": "2026-02-01T12:00:00.000Z",
@@ -62,18 +64,21 @@ node scripts/tool-whitelist/test-models-tools.mjs --provider openai
 ```
 
 **Options**:
+
 - `--limit N` - Test only first N models per provider
 - `--provider name` - Test only specific provider (openai|anthropic|gemini|openrouter)
 
 **Output**: `scripts/tool-whitelist/tool-test-results.json`
 
 **What it does**:
+
 - Loads models from `available-models.json`
 - Makes actual API calls with tool definitions
 - Records which models successfully call tools
 - Saves detailed results including errors
 
 **Example output**:
+
 ```json
 {
   "testedAt": "2026-02-01T12:30:00.000Z",
@@ -107,19 +112,22 @@ node scripts/tool-whitelist/generate-whitelist.mjs --output my-whitelist.txt
 ```
 
 **Options**:
+
 - `--min-success-rate N` - Minimum success rate 0-1 (default: 1.0)
 - `--output path` - Custom output path (default: scripts/tool-whitelist/generated-whitelist.txt)
 
 **Output**: `scripts/tool-whitelist/generated-whitelist.txt`
 
 **What it does**:
+
 - Reads `tool-test-results.json`
 - Filters models that support tools
 - Groups by base pattern (removes date suffixes)
 - Generates formatted whitelist
 - Shows comparison with current whitelist
 
 **Example output**:
+
 ```
 # OpenAI
 gpt-5.2
@@ -143,11 +151,11 @@ After reviewing the generated whitelist:
 
 ## Files Generated
 
-| File | Purpose | Created By |
-|------|---------|------------|
-| `available-models.json` | All available models from providers | Step 1 |
-| `tool-test-results.json` | Tool support test results | Step 2 |
-| `generated-whitelist.txt` | New default whitelist | Step 3 |
+| File                      | Purpose                             | Created By |
+| ------------------------- | ----------------------------------- | ---------- |
+| `available-models.json`   | All available models from providers | Step 1     |
+| `tool-test-results.json`  | Tool support test results           | Step 2     |
+| `generated-whitelist.txt` | New default whitelist               | Step 3     |
 
 ## Example Full Workflow
 
@@ -177,23 +185,27 @@ cat scripts/tool-whitelist/generated-whitelist.txt
 ### Testing Strategy
 
 **Quick validation** (fast, cheap):
+
 ```bash
 node scripts/tool-whitelist/test-models-tools.mjs --limit 5
 ```
 
 **Thorough validation** (slow, expensive):
+
 ```bash
 node scripts/tool-whitelist/test-models-tools.mjs
 ```
 
 **Provider-specific**:
+
 ```bash
 node scripts/tool-whitelist/test-models-tools.mjs --provider gemini
 ```
 
 ### Automation
 
 These scripts are **intentionally manual**:
+
 - No automatic updates to source code
 - Human review required
 - Prevents accidental whitelist corruption
@@ -203,13 +215,15 @@ These scripts are **intentionally manual**:
 ### "Error loading available-models.json"
 
 Run step 1 first:
+
 ```bash
 node scripts/tool-whitelist/fetch-available-models.mjs
 ```
 
 ### "Error loading tool-test-results.json"
 
 Run step 2 first:
+
 ```bash
 node scripts/tool-whitelist/test-models-tools.mjs
 ```
@@ -223,6 +237,7 @@ node scripts/tool-whitelist/test-models-tools.mjs
 ## Maintenance Schedule
 
 **Recommended frequency**:
+
 - **Monthly**: Quick validation (`--limit 5`)
 - **Quarterly**: Full validation (all models)
 - **On-demand**: When new major models are released