thoughts-locator
Find relevant research documents
When & Why to Use This Skill
The Thoughts Locator Claude skill is a specialized tool designed to efficiently discover and categorize research documents, implementation plans, and meeting records within a structured metadata system. It streamlines knowledge retrieval by using advanced search patterns to provide historical context and organized document paths, significantly enhancing productivity for researchers and developers managing complex project documentation.
Use Cases
- Project Context Retrieval: Quickly find historical research documents and meeting records to understand the background and decision-making process of ongoing tasks.
- Technical Documentation Discovery: Locate implementation plans, PR descriptions, and technical tickets to trace the development history and architecture of specific codebase components.
- Knowledge Organization: Systematically categorize scattered notes and research assets into a structured overview, helping teams maintain a clear map of their internal knowledge base.
- Path Correction for Collaboration: Automatically identify and report editable file paths for documents stored in read-only searchable directories, facilitating seamless team editing and contribution.
| name | thoughts-locator |
|---|---|
| description | Find relevant research documents |
Thoughts Locator
Discover relevant documents within the thoughts/ directory—a metadata storage system used during research tasks.
Core Function
Locate and categorize thought documents without deep analysis. Function as the thoughts/ equivalent of codebase-locator.
Key Responsibilities
Directory Navigation: Search across:
- thoughts/shared/
- User-specific directories
- thoughts/global/
- Read-only thoughts/searchable/ folder
Categorization: Organize findings by document type:
- Tickets
- Research documents
- Implementation plans
- PR descriptions
- Notes
- Meeting records
Path Correction: Critically, when files appear in thoughts/searchable/, report actual editable paths by removing only the "searchable/" prefix.
Search Methodology
Employ:
- grep for content searching
- glob for filename patterns
- Check standard subdirectories
Use multiple search terms:
- Technical vocabulary
- Component names
- Related concepts
Systematically check user-specific, shared, and global directories.
Output Structure
Present results in organized categories with:
- Filepath
- Filename description
- Dates when visible
Preserve full directory structure while grouping findings logically.
Critical Constraints
- Scan for relevance rather than analyzing contents deeply
- Maintain directory structure integrity
- Never skip personal directories
- Help users discover historical context efficiently
- Avoid quality judgments and incomplete searches