Skip to content

Releases: bkataru/zeitgeist

v1.1.1 - mcp.zig Update

26 Jan 11:39

Choose a tag to compare

What's Changed

Changed

  • Updated mcp.zig dependency to v0.2.1 with protocol version fix (2024-11-05) and memory leak fix
    • Fixes protocol version from invalid "2025-11-25" to correct "2024-11-05"
    • Fixes ResourceRegistry.list() memory leak with new freeList() method

Full Changelog: v1.1.0...v1.1.1

v1.1.0 - MCP Resources & Progress Notifications

26 Jan 11:22

Choose a tag to compare

What's New

MCP Server Enhancements

This release significantly enhances the MCP (Model Context Protocol) server with new capabilities for AI agents and LLM tools.

New Resources

  • zeitgeist://cwd - Returns the current working directory path
  • zeitgeist://tree - Returns a directory tree listing of the current directory
  • zeitgeist://file/{path} - Returns the content of a specific file

Progress Notifications

Tool handlers now send real-time progress updates during pack/scan operations, allowing clients to display progress indicators.

Cancellation Support

Long-running operations can now be cancelled mid-execution via MCP cancellation tokens.

Bug Fixes

  • Fixed memory leaks in resources/list and resources/read MCP handlers using arena allocators
  • Fixed protocol version in mcp.zig dependency from invalid "2025-11-25" to correct "2024-11-05"

Dependencies

  • Updated mcp.zig to v0.2.0

Documentation

  • Added comprehensive MCP Server documentation to API.md
  • Updated README with MCP resources section

Full Changelog: v1.0.0...v1.1.0

Zeitgeist v1.0.0

26 Jan 02:33

Choose a tag to compare

Zeitgeist v1.0.0 - Initial Release

Zeitgeist (/ˈtsaɪtɡaɪst/) — German for "spirit of the age"
Capture the essence of your codebase for LLM consumption.

Highlights

  • Multiple Output Formats: XML, Markdown, JSON, and plain text
  • Smart Filtering: Glob patterns with .gitignore and .ignore support
  • Token Counting: Accurate estimation for LLM context planning
  • Output Chunking: Split large outputs with --max-tokens and --split
  • AI Skill Output: Generate structured documentation with --skill-output
  • Git Integration: Sort by recency, include diffs and logs
  • Remote Repository Support: Process GitHub/GitLab/Bitbucket repos directly
  • Security Scanning: Detect and redact secrets with --security-check
  • Shell Completions: Bash, Zsh, and Fish completion scripts
  • Man Page: Auto-generated documentation

Installation

# Build from source (requires Zig 0.14+)
zig build -Doptimize=ReleaseFast

# Or download pre-built binaries from releases

Quick Start

# Pack current directory
zg .

# Pack with token limit
zg . --max-tokens 128k

# Generate AI skill documentation
zg . --skill-output docs/

# Process remote repository
zg https://github.com/user/repo

See README.md for full documentation.