Context Harness

Standalone context harness library for managing chat thread history, messages, and context construction for AI applications and agents.

Features

• Thread Management: Conversation thread lifecycle with message history
• Memory Management: Cross-thread memories with explicit traceability
• Experience Distillation: Extract summaries and reusable patterns
• Context Construction: Intelligent selection and formatting for new queries
• Message Spilling: Automatic offloading of large content with insertion syntax markers
• LangChain Integration: Seamless compatibility with LangChain message types

Installation

pip install context-harness

Quick Start

from langchain_core.messages import HumanMessage, AIMessage
from context_harness import ContextHarness

# Initialize harness (uses FileBackend by default)
harness = ContextHarness()

# Create thread
thread_id = harness.create_thread(metadata={"user_id": "user123"})

# Add messages
harness.add_message(thread_id, HumanMessage("Hello"))
harness.add_message(thread_id, AIMessage("Hi! How can I help?"))

# Get history
history = harness.get_thread_history(thread_id)

# Build context for new query
context = harness.build_context_for_query("What's the pricing?", thread_id)

Architecture

Three-layer architecture:

• Interface Layer: ContextHarness facade for simple API
• Core Layer: Domain components (ChatHistoryManager, MemoryManager, ExperienceDistiller, ContextBuilder)
• Data Layer: StorageBackend abstraction with FileBackend default

Documentation

See docs/specs/ for RFC specifications and docs/impl/ for implementation guides.

License

MIT License - see LICENSE file for details.

Development

Install development dependencies:

pip install -e ".[dev]"

Run tests:

pytest tests/

Status

• ✅ RFC-001: World View (Conceptual Design)
• ✅ RFC-002: Core Architecture (Architecture Design)
• ✅ RFC-003: Domain Models (Implementation Interface Design) - IMPLEMENTED
• 🔄 RFC-004: Storage Backend (Implementation Interface Design)
• 🔄 RFC-005: Facade Interface (Implementation Interface Design) Harnessing agentic context