obsidian-second-brain

Indexes the entire Obsidian vault as a searchable knowledge base, enabling Claude to retrieve relevant notes based on semantic similarity rather than keyword matching. Uses embeddings to understand context and relationships between notes, allowing the agent to surface connected information across the vault without explicit linking. Implements local indexing to avoid sending vault contents to external services.

Enables creation of background agents that run on a schedule (hourly, daily, weekly) to perform research tasks, synthesize information, and update notes without manual intervention. Agents execute Claude Code skill commands in sequence, reading from the vault, processing information, and writing results back to specified notes. Implements a scheduling system that persists agent configurations and execution history.

Provides a unified interface for executing 31+ Claude Code skill commands that manipulate vault content, including note creation, editing, searching, and analysis. Implements a command registry that maps natural language requests to specific commands, handles parameter binding, and manages execution context. Supports command chaining and conditional execution based on results.

Analyzes vault content to identify patterns, trends, gaps, and insights. The agent can identify frequently discussed topics, track how concepts evolve across notes, identify knowledge gaps, and generate insights about the vault's content. Supports statistical analysis and visualization data generation for vault structure and content patterns.

Imports notes from external sources (markdown files, web content, PDFs, other note-taking apps) and normalizes them into Obsidian-compatible format with consistent metadata and structure. The agent parses various formats, extracts content and metadata, and generates Obsidian-compatible markdown with appropriate frontmatter, links, and tags. Supports batch import with deduplication.

Provides writing assistance and editing capabilities that are aware of vault content and style. When editing notes, the agent can suggest improvements, check consistency with vault conventions, identify redundancy with existing notes, and improve clarity while maintaining the user's voice. Supports style checking and tone analysis based on vault examples.

Chains multiple Claude Code skill commands together to perform complex transformations on vault content, such as bulk note reformatting, metadata extraction, or content reorganization. Implements a pipeline abstraction that passes output from one step as input to the next, with error handling and rollback capabilities. Supports conditional branching based on note properties or content analysis.

Generates new notes or expands existing ones based on vault context, using semantic search to pull relevant information and Claude to synthesize new content. When creating a note, the agent retrieves related notes from the vault, uses them as context, and generates content that integrates with existing knowledge. Supports templates and structured generation for consistent note formats.

Analyzes note content to identify semantic relationships and automatically creates or suggests links between notes. Uses Claude to understand note semantics and determine when notes should be linked, then updates Obsidian's link graph. Supports bidirectional link creation and can detect and resolve duplicate or conflicting links.

Generates code snippets, functions, or documentation based on patterns and examples found in the vault. When a user requests code generation, the agent searches the vault for relevant examples, coding patterns, or documentation, then uses those as context to generate code that matches the vault's conventions and style. Supports multiple programming languages and documentation formats.

Automatically synthesizes information from multiple notes into coherent research summaries, literature reviews, or analysis documents. The agent retrieves relevant notes based on a research query, analyzes them for key findings and themes, and generates a synthesized document that integrates insights across sources. Supports citation tracking and source attribution.

Analyzes unstructured note content to extract structured metadata (entities, relationships, properties) and populates note frontmatter with extracted data. Uses Claude to understand note content and identify relevant metadata, then updates YAML frontmatter with extracted information. Supports custom metadata schemas and validation rules.

Generates or expands notes in response to natural language queries, using vault context to ensure generated content is relevant and integrated. When a user asks a question, the agent searches the vault for relevant information, identifies gaps, and generates new content or expands existing notes to answer the question comprehensively. Supports follow-up queries and iterative refinement.

Integrates task and project management capabilities with vault content, enabling creation of tasks from notes, tracking project progress, and generating project summaries. The agent can create task notes with metadata, track dependencies between tasks, and generate project status reports by analyzing task notes. Supports integration with Obsidian's task syntax and metadata.