Roadmap

ck’s development roadmap, completed features, and planned improvements.

Current Version (v0.5.3)

Completed Features ✅

Core Search

• grep-compatible CLI with semantic search capabilities
• Semantic search using local embedding models
• Hybrid search combining semantic + keyword with Reciprocal Rank Fusion
• File listing flags (-l, -L) for grep compatibility
• Threshold filtering for relevance control
• Relevance scoring with visual highlighting
• Complete code section extraction (–full-section flag)

AI Integration

• MCP (Model Context Protocol) server for AI agent integration
• Built-in pagination for large result sets
• JSONL output format for streaming workflows
• Structured output (JSON/JSONL) for automation

Indexing & Performance

• Incremental index updates with hash-based change detection
• Automatic delta indexing (only changed files)
• Smart file filtering with .gitignore and .ckignore support
• Clean stdout/stderr separation for reliable scripting
• Interrupt handling (safe Ctrl+C during indexing)

Language Support

• Tree-sitter parsing for 8 languages (Python, JS/TS, Rust, Go, Ruby, Haskell, C#, Zig)
• Intelligent chunking at function/class boundaries
• Token-aware chunking with HuggingFace tokenizers
• Smart binary detection (content-based, not extension-based)

Embedding Models

• FastEmbed integration with multiple models
• BGE-Small (default, fast, 400-token chunks)
• Nomic V1.5 (8K context, 1024-token chunks)
• Jina Code (code-specialized, 8K context)
• Model switching with automatic rebuild detection

User Experience

• File exclusion patterns and glob support
• .ckignore file with automatic creation and sensible defaults
• Published to crates.io (cargo install ck-search)
• Cross-platform support (Linux, macOS, Windows)
• Comprehensive documentation with examples

Next Release (v0.6)

Planned Features 🚧

Configuration

• Configuration file support (.ck.toml or ck.config.json) — Default model selection — Default threshold and top-k values — Per-project preferences — Custom exclusion patterns

Distribution

• Package manager distributions — Homebrew formula for macOS — APT packages for Debian/Ubuntu — RPM packages for Fedora/RHEL — Chocolatey package for Windows — Pre-built binaries for all platforms

MCP Enhancements

• Enhanced MCP tools — File writing capabilities — Refactoring assistance — Code transformation tools — Bulk operations

Editor Integration

• VS Code extension — Inline semantic search — Code navigation — Quick actions — Status bar integration

• JetBrains plugin — IntelliJ IDEA support — PyCharm support — WebStorm support — Common features across IDEs

Language Support

• Additional language chunkers — Java (with tree-sitter-java) — PHP (with tree-sitter-php) — Swift (with tree-sitter-swift) — Kotlin (requested #21) — C/C++ (requested #21)

Future Considerations

Community Requests

Based on GitHub issues and discussions:

File Type Filtering (#28)

Status: Requested Description: Add --type flag similar to ripgrep

# Proposed syntax
ck --type rust "pattern"
ck --type js,ts "pattern"
ck -T rust "pattern"  # Short form

Use case: Quickly filter searches to specific file types without glob patterns.

Git Diff-Based Re-indexing (#69)

Status: Under exploration Description: Use git diff for incremental updates of large files

Challenge: Semantic chunking boundaries can change based on file content, making diff-based updates complex.

Potential approach:

• Implement for size-based chunking first
• Explore heuristics for semantic chunking
• May require trade-offs in accuracy vs speed

External Embedding API Support (#49)

Status: Under consideration Description: Support external embedding services

Requested APIs:

• OpenAI embeddings
• HuggingFace Inference API
• Anthropic Claude embeddings
• Custom HTTP endpoints

Trade-offs:

• ✅ Access to latest/proprietary models
• ✅ No local model storage
• ❌ Requires internet connection
• ❌ Privacy/security considerations
• ❌ API costs

Bug Detection Category (#23)

Status: Research phase Description: Specialized search for potential bugs

Possible implementation:

• Pre-trained model for code smell detection
• Pattern-based heuristics
• Integration with static analysis tools
• Custom queries for common bug patterns

Long-Term Vision

Performance

• GPU acceleration for embedding generation
• Distributed indexing for monorepos
• Caching strategies for frequently searched patterns
• Streaming search results

Advanced Features

• Code similarity detection (find duplicate code)
• Dependency graph analysis
• Cross-repository search
• Code metrics and quality analysis
• Integration with code review tools

AI Capabilities

• Context-aware code completion
• Automated documentation generation
• Code explanation and summarization
• Refactoring suggestions
• Test generation

Enterprise Features

• Team collaboration features
• Shared index repositories
• Access control and permissions
• Audit logging
• On-premise deployment options

Contributing to the Roadmap

How to Influence Priority

1. Vote on existing issues: Use 👍 reactions on GitHub issues
2. Open feature requests: Describe your use case clearly
3. Contribute code: Submit PRs for features you need
4. Sponsor development: Support prioritization of specific features

Feature Request Guidelines

When requesting features:

• ✅ Describe the problem, not just the solution
• ✅ Provide concrete use cases
• ✅ Consider implementation complexity
• ✅ Think about API design
• ✅ Check if workarounds exist

Good example:

“As a developer working on microservices, I need to search across multiple repos simultaneously. Currently I have to cd into each repo and search separately, which is tedious for finding related implementations.”

Less helpful:

“Add multi-repo search”

Version Timeline

Version	Focus	Status
v0.1	MVP, basic search	✅ Released
v0.2	Tree-sitter, chunking	✅ Released
v0.3	Incremental indexing	✅ Released
v0.4	Multiple models	✅ Released
v0.5	MCP integration	✅ Released (current)
v0.6	Config, distribution	🚧 In planning
v0.7	Editor integrations	📋 Planned
v0.8	Advanced features	💭 Conceptual

Breaking Changes Policy

ck follows Semantic Versioning:

• MAJOR (1.0.0): Breaking changes to CLI or API
• MINOR (0.X.0): New features, backward compatible
• PATCH (0.0.X): Bug fixes, backward compatible

v1.0 Stability

Before reaching v1.0, ck will:

• Stabilize CLI interface
• Finalize MCP tool signatures
• Complete core feature set
• Achieve production maturity
• Document upgrade paths

See Also

• Known Limitations — Current constraints
• FAQ — Common questions
• Contributing Guide — How to contribute
• GitHub Issues — Feature requests and bugs
• Changelog — Detailed release history