Settings Reference
All settings are prefixed with ollama-code-review.*.
| Setting | Description | Default |
|---|---|---|
model | AI model to use (Ollama, Claude, Gemini, etc.) | custom |
autoSelectModel | Automatically switch to the best model based on task | false |
cache.enabled | Reuse exact review results when prompt and model config match | true |
cache.ttlMinutes | Minutes before cached review results expire | 1440 |
cache.maxEntries | Maximum review results to keep in local cache | 50 |
claudeApiKey | Anthropic API key | "" |
geminiApiKey | Google AI Studio API key | "" |
mistralApiKey | Mistral AI API key | "" |
hfApiKey | Hugging Face API token | "" |
openaiCompatible.endpoint | Base URL for OpenAI-compatible server | http://localhost:1234/v1 |
endpoint | Local Ollama instance generate API endpoint | http://localhost:11434/api/generate |
temperature | Creativity of the AI response (0.0 to 1.0) | 0 |
frameworks | Target frameworks (e.g., React, Node.js) | ["React"] |
diffFilter.ignorePaths | Glob patterns for paths to ignore | node_modules, ... |
preCommitGuard.severityThreshold | Severity level to block commits | high |
contextGathering.enabled | Include related files as context | true |
notifications.slack.webhookUrl | Slack webhook URL | "" |
knowledgeBase.enabled | Enable Team Knowledge Base | true |
rag.enabled | Enable RAG context injection | false |
streaming.enabled | Enable token-by-token output | true |
mcp.enabled | Enable built-in MCP server | false |
mcp.port | Port for the built-in MCP server | 19840 |
mcp.autoKillPortConflicts | Automatically kill processes using the MCP port | false |
mcp.allowedOrigins | Allowed origins for browser MCP clients | ["chrome-extension://*"] |
mcp.authToken | Security token for browser MCP clients | "" |
mcp.semble.pythonPath | Python executable for Semble-backed MCP code search. Leave empty to use python3 on macOS/Linux or python on Windows. | "" |
mcp.externalServers | Configuration for connecting to external MCP servers | {} |
autoReview.enabled | Enable background review on file save | false |
autoReview.minSeverity | Only notify for findings at or above this level | high |
coverage.includeGlob | Glob pattern for files to include in the Coverage view | **/*.{ts,js,tsx,jsx,py,java,cs,go,rb,php,rs,swift,kt,vue,svelte,c,cpp,h} |
coverage.excludeGlob | Comma-separated glob patterns to exclude from Coverage | node_modules, dist, build, ... |
coverage.staleAfterDays | Mark files as stale after this many days since their last review | 14 |
coverage.maxFiles | Maximum workspace files to scan for Coverage | 1000 |
contentstack.enabled | Enable Contentstack schema validation | false |
contentstack.schemaSource | Source for schemas (local or api) | local |
copyFunction.maxDepth | BFS depth for call-graph expansion in Copy Function with Imports | 3 |
copyFunction.maxFunctions | Max functions to collect in Copy Function with Imports | 15 |
copyFunction.maxCharsPerFunction | Max characters per function body in Copy Function with Imports | 8000 |
copyFunction.characterBudget | Total character budget for all collected functions | 64000 |
copyWithImports.noLimits | Remove all Smart Context limits (use carefully on large files) | false |
restoreLastReview | Auto-restore the last review's findings and annotations on startup | true |
codeHealth.enabled | Enable per-file health tracking and regression detection | true |
codeHealth.regressionThreshold | Score drop (points) that triggers a regression warning | 10 |
codeHealth.blockOnRegression | Block commits when a file health regression is detected | false |
codeHealth.hotspotCount | Number of worst-scoring files to display in the Code Health view | 15 |
Exhaustive List
For the full list of over 50 configuration options, please refer to the VS Code Settings UI by searching for @ext:VinhNguyen-Vincent.ollama-code-review.