Agentic Coding Harness

An agentic coding harness optimized for Z.AI GLM-4.7 models.

Features

• Tool Calling: Reliable function calling with proper error handling
• File Operations: Read, write, and edit files with precision
• Code Editing: String-replace based editing with context matching
• File Search: Glob-based file discovery
• Shell Commands: Execute shell commands with output/error capture
• Indexing: Automatic code indexing for search
• Lexical Search: Fast text/regex search across codebase
• Semantic Search: Natural language code search using embeddings
• Context Compaction: Automatic context management to stay within limits

Installation

cd harness
pip install -e ".[dev]"

Configuration

Create a .env file:

LLM_API_URL=https://api.z.ai/api/paas/v4/
LLM_API_KEY=your-api-key
LLM_MODEL=glm-4.7

Usage

Interactive Mode

python -m harness.cli

Single Command

python -m harness.cli -m "Read the README.md and summarize it"

Options

• -e, --env: Path to .env file (default: .env)
• -w, --workspace: Workspace directory (default: current)
• -m, --message: Single message to run
• --max-iterations: Maximum iterations (default: 20)
• --max-tokens: Maximum context tokens (default: 32000)

Running Tests

# Run all unit tests
pytest

# Run with coverage
pytest --cov=harness

# Run integration tests (requires API key)
pytest -m integration

# Run specific test file
pytest tests/test_file_tools.py

Architecture

src/harness/
├── __init__.py      # Package exports
├── agent.py         # Main Agent class
├── cli.py           # Command-line interface
├── config.py        # Configuration management
├── context.py       # Context window management
├── llm_client.py    # LLM API client
└── tools/
    ├── __init__.py
    ├── registry.py      # Tool registry
    ├── file_tools.py    # File operations
    ├── shell_tools.py   # Shell commands
    └── search_tools.py  # Search functionality

Tools Available

Tool	Description
read_file	Read file content, optionally with line ranges
write_file	Write content to a file
edit_file	Replace a specific string in a file
list_directory	List directory contents
file_search	Search for files by glob pattern
run_command	Execute shell commands
lexical_search	Text/regex search in files
semantic_search	Natural language code search

License

MIT