knowledge

You can retrieve project coding standards and architecture documentation by applying hybrid link-graph, vector, and keyword search across docs, references, and notes to obtain authoritative context for LLM guidance.

The best way to search project knowledge is using a unified hybrid retrieval system that combines link-graph traversal, vector search, and keyword matching across ingested documents and durable notes for high-precision results.

You ingest documents by chunking and embedding content, extracting entities into a knowledge graph, and summarizing session trajectories to maintain long-term memory and project context.

Yes, you can load development context, language-specific coding standards, and writing memory to enforce project conventions before editing or committing code changes.

Yes, the knowledge retrieval tool exposes skill command endpoints for status checks, index refresh, dependency symbol search, ingest operations, and knowledge-base maintenance.

You need to summarize session trajectories for long-term memory during development and documentation tasks when you require durable notes and contextual continuity across complex project workflows.