Skill v1.0.1
currentAutomated scan100/100+1 new
version: "1.0.1" name: Research Knowledge Base description: Maintain a searchable knowledge base for a research topic. allowed-tools:
- knowledge_search
- knowledge_read
- create_note
- update_note
- context_search
- context_read
- search_items
- search_fulltext
- search_notes
- search_by_tag
- get_item
- get_item_metadata
- get_item_notes
- get_note_content
- get_collections
- get_collection_items
- get_related_items
- get_recent
- get_outline
- list_sections
- get_paper_section
- search_paper_content
- get_annotations
- get_paper_metadata
- open_item
triggers:
- research knowledge base
- knowledge base
- 研究知识库
- 知识库
- 课题追踪
Maintain continuity in the user's research through the unified knowledge index.
Start with knowledge_search, then knowledge_read for relevant topics, open questions and Zotero notes. Follow the returned source references. Read the current note before proposing an update; reuse its libraryID and key. Save documents as Zotero notes using the existing approval flow. Never ask the user to create a knowledge base, choose a topic container or supply an internal knowledge-base ID.
Research memory is maintained automatically after ordinary conversations when enabled. Describe the user's explicit goal, current supported findings, unresolved/resolved/dropped questions, and next action accurately in your answer. Reuse an existing topic when the scope matches. Do not infer a permanent interest from a one-off paper question or claim that automatic maintenance succeeded without a receipt.
Separate evidence from interpretation and link source papers as libraryID:key. Use context tools for working progress. User-corrected and protected memories take precedence; current instructions override older preferences. Legacy knowledge files remain readable sources, not new write destinations.