Skip to content

Configuration

Vault Intelligence is designed to be powerful out of the box, but you can customise it to suit different hardware, budgets, and workflows.

Connection

Manage your core intelligence providers and credentials.

SettingDefaultDescription
Google API keyNoneYour secret key from Google AI Studio. Stored securely in your system keychain (or masked in data.json if secure storage is unavailable). Required for all Gemini models and embeddings.
Ollama endpointhttp://localhost:11434Server URL for local model providers. Required for Ollama-based chat and embeddings.
Refresh model listNoneA manual trigger to fetch available models from the Gemini API and local Ollama server.

Researcher

Customise your primary research assistant’s intelligence and writing style.

SettingDefaultDescription
Chat modelgemini-3-flash-previewThe primary reasoning engine.
Cloud (Gemini): High quality, zero local overhead.
Local (Ollama): Fully private and free to run.
Agent languageEnglish (US)The primary language for agent responses. This affects the default system prompt and output style.
System instructionDefault (Managed)The core personality and rules for the Researcher. Leave as default to receive automatic improvements in future updates.
Context window budget200,000 tokensThe maximum volume of notes the AI can "read" at once. Higher budgets allow for broader research but may hit API limits.

> [!TIP]
> If you experience "429 Too Many Requests" errors, try lowering this budget. See the [Troubleshooting Guide](file:///home/allane/Code/GitHub/obsidian-vault-intelligence/docs/reference/troubleshooting.md#429-too-many-requests) for details.
Max agent steps5Limits the number of "reasoning loops" the agent can take to prevent infinite thinking or excessive API costs.
Author nameMeThe name used when the agent refers to you or credits content. This also acts as a fallback for missing author metadata in your notes.
Context-aware headerstitle, tags...A list of frontmatter properties the AI should always be aware of when analyzing a note chunk.
Enable web searchOnAllows the agent to verify facts and fetch live news from the internet.
Web search modelgemini-2.5-flash-liteA cost-effective model specialised for verifying information and searching the web.
Enable link contextOnAllows Gemini 3.1+ models to natively analyse URLs using Google's internal retrieval system for higher accuracy.
Enable computational solverOnAllows the agent to write and execute Python code for complex math and data analysis.
Code execution modelgemini-3-flash-previewThe specialized model used for generating Python code.
Enable agent write accessOffAllows the agent to create or update notes. Security Note: This always requires manual confirmation before any file is changed.
Vault reading limit25Maximum number of notes the researcher can retrieve to answer a single question.

Explorer

Fine-tune how connections are discovered and how the semantic graph is visualised.

Embedding Engine

SettingDefaultDescription
Embedding providergeminiGoogle Gemini: Cloud-based (requires API key).
Ollama: Local server (requires Ollama).
Transformers.js: 100% local CPU processing (no server required).
Embedding modelgemini-embedding-001The vector engine used to calculate relationships. Local models (eg nomic-embed-text) are downloaded once (~25MB--150MB).
Embedding dimension768Output vector size. Higher dimensions provide better accuracy but result in a larger search index on disk.
Embedding chunk size1024 / 512Character count per vector segment. Automatically adjusts to 512 for complex scripts (Chinese, Japanese) or local models to improve retrieval quality.

Search Strategy

SettingDefaultDescription
Enable dual-loop searchOnCombines fast local vector search (Loop 1) with deep AI re-ranking (Loop 2). This significantly reduces "hallucinations" in search results.
Re-ranking modelgemini-3-flash-previewThe AI engine used by the "Analyst" (Loop 2) to verify and rank the most relevant notes.
Minimum similarity score0.5Relevance threshold (0.0 to 1.0). Lower this if search results feel too sparse.
Keyword match weight1.2Calibration for keyword vs vector search. Higher values make keyword matches more conservative in hybrid results.
Similar notes limit20Max results displayed in the sidebar when looking for related content.

TIP

If your search results are missing known information, consult the [Search Quality](file:///home/allane/Code/GitHub/obsidian-vault-intelligence/docs/reference/troubleshooting.md#the-agent-says-it-cant-find-information-but-i-know-i-have-a-note-on-it) section of the troubleshooting guide.

Semantic Galaxy

SettingDefaultDescription
Semantic graph node limit250Maximum number of nodes rendered in the Galaxy view and search expansion.
Structural edge thickness1.0Visual weight of your explicit [[wikilinks]].
Semantic edge thickness0.5Visual weight of implied relationships discovered by AI.
Implicit folder semanticsontologyControls how your folder structure is weighted.
none: Folders are ignored.
ontology: Folders act as topics only if they match your fixed Ontology.
all: Every folder is treated as a semantic topic.

Gardener

Configure the automated hygiene agent to maintain your vault's ontology.

SettingDefaultDescription
Gardener modelgemini-3-flash-previewThe model used for structural analysis and improvement suggestions.
Gardener rulesDefault (Managed)The persona and rules for the Gardener. Leave as default to receive automatic updates.
Ontology pathOntologyFolder where concept, entity, and MOC (Map of Content) notes are stored.
Semantic merge threshold0.85Similarity score required to suggest merging two similar topics. Set to 1.0 to disable merging.

Orphan Management

SettingDefaultDescription
Orphan grace period7 daysNumber of days a note must be unlinked/orphaned before the Gardener suggests pruning it.
Archive folder pathOntology/_ArchiveWhere pruned or "deleted" notes are moved by the Gardener for safekeeping.
Gardener plans pathGardener/PlansFolder where proposed hygiene plans are saved for your review.
Plans retention7 daysDuration to keep old plan files before they are automatically purged.

Advanced Systems

Technical tuning for performance and security.

SettingDefaultDescription
Indexing delay5000msWait time after typing stops before re-indexing in the background.
Indexing throttle100msTechnical delay between files during processing to avoid API rate limiting.
Search centrality limit50Max number of "bridge" nodes pulled from the graph to expand search context.
Allow local network accessOffAdvanced: Allows the agent to access localhost or private network IPs.
> [!CAUTION]
> Enabling this makes you vulnerable to SSRF (Server-Side Request Forgery) attacks. Only enable if you are running local tools you fully trust.
Log levelWarnConsole verbosity. Set to Debug when collecting information for bug reports.

Privacy and Storage

Vault Intelligence stores its search index and relationship graph in a specialised binary format inside the plugin's data/ directory.

  • Automated .gitignore: The plugin automatically manages a .gitignore file for its data folder to prevent massive index files from bloating your Git repository or causing sync conflicts.
  • Data Safety: The index is a derived cache; it can be regenerated from your notes at any time by clicking Re-index Vault in the Explorer settings.
  • Secure Storage: On supported systems (macOS, Windows, Linux with libsecret), your API keys are stored in the OS keychain, never in plain text.