Skip to content

CLAUDE.md

Latest commit

 

History

History
50 lines (43 loc) · 3.24 KB

CLAUDE.md

File metadata and controls

50 lines (43 loc) · 3.24 KB

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Build & Run Commands

Only apply these to production code, not testing scripts. Don't run this after a code change, only run if explicitly asked by user.

  • Lint: ruff check --fix --line-length=120 --select=I <path>
  • Format: ruff format --line-length=120 <path>
  • Install dependencies with uv: uv add <package_name>
  • Always use uv to run Python

Heurist Mesh Overview

  • mesh/: Heurist Mesh specialized agent implementations Heurist Mesh is an open network of modular and purpose-built AI agents. Technically, you can think of each agent as a collection of tools for AI. Once a Mesh agent is added to the main branch, it's automatically deployed and instantly available via REST API and MCP (Model Context Protocol).

Mesh Agent Development Guidelines

  • Base class: Inherit from mesh.mesh_agent.MeshAgent
  • Agent naming: Use descriptive PascalCase names ending with "Agent" (e.g., AlloraPricePredictionAgent)
  • Metadata: Include complete metadata in __init__ (name, version, author, description, inputs, outputs, etc.)
  • API access: Store sensitive API keys in environment variables, never hardcode
  • Testing: Create a test file in mesh/tests/ with example input/output saved as YAML
  • Tools: Define tools using get_tool_schemas() method
  • External APIs: Use @with_retry and @with_cache decorators for network operations
  • Resource management: Implement __aenter__ and __aexit__ for proper cleanup
  • Documentation: Include examples and clear descriptions in metadata

Testing Mesh Agents and Modules

When developing or testing individual Mesh agent modules to see outputs ad-hoc, create test scripts in mesh/test_scripts/ folder:

  • Always use uv run: Run scripts with uv run python mesh/test_scripts/<script_name>.py
  • Test script location: Place all test scripts in mesh/test_scripts/ directory
  • Import path fix: Add project root to Python path in test scripts: 
    import sys
from pathlib import Path
project_root = Path(__file__).parent.parent.parent
sys.path.insert(0, str(project_root))
  • Test specific functionality: Test individual agent tools, API responses, or data transformations
  • Example tests: See mesh/test_scripts/test_trending_token.py for aggregated data testing, or test_token_resolver.py for individual agent testing

Coding Style

  • Prefer the most straightforward way of implementation.
  • No defensive programming. Use try/except sparingly. Let most fatal errors propogate.
  • No adding extra variables if there's not a clear reason to do so. Prefer simplicity
  • Use comments sparingly. Function and variable names should be self-explaining and you don't need to comment most of the times
  • If duplicate code is written, consider refactoring.
  • DO NOT use getattr or isinstance or defensive programming patterns
  • Do not create summary doc after completing a task, unless explicitly asked
  • When you are asked to review git staged code changes, give feedback as if you are a senior tech lead ensuring high standards and rank the issues found with high/medium/low. Do not edit code or run git commands that manipulate the repo.