llama-farm/llamafarm / .claude/skills/common-skills
common-skills
VerifiedBest practices for the Common utilities package in LlamaFarm. Covers HuggingFace Hub integration, GGUF model management, and shared utilities.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.5 KB |
| huggingface.md | markdown | 12.1 KB |
SKILL.md
View on GitHub---
name: common-skills
description: Best practices for the Common utilities package in LlamaFarm. Covers HuggingFace Hub integration, GGUF model management, and shared utilities.
allowed-tools: Read, Grep, Glob
user-invocable: false
---
# Common Skills for LlamaFarm
Best practices and code review checklists for the `common/` package - shared Python utilities used across all LlamaFarm services.
## Component Overview
| Attribute | Value |
|-----------|-------|
| Path | `common/` |
| Package | `llamafarm-common` |
| Python | 3.10+ |
| Key Dependencies | huggingface_hub, hf-transfer |
## Purpose
The `common/` package provides shared functionality that needs to be consistent across multiple Python services:
- Model file utilities (GGUF selection, quantization parsing)
- HuggingFace Hub integration (listing, downloading)
- Process management (PID files)
## Shared Python Skills
This skill inherits all patterns from the shared Python skills:
| Topic | File | Relevance |
|-------|------|-----------|
| Patterns | [../python-skills/patterns.md](../python-skills/patterns.md) | Dataclasses, type hints, comprehensions |
| Typing | [../python-skills/typing.md](../python-skills/typing.md) | Type annotations, modern syntax |
| Testing | [../python-skills/testing.md](../python-skills/testing.md) | Pytest fixtures, mocking HuggingFace APIs |
| Errors | [../python-skills/error-handling.md](../python-skills/error-handling.md) | Custom exceptions, logging |
| Security | [../python-skills/security.md](../python-skills/security.md) | Path validation, safe file handling |
## Framework-Specific Checklists
| Topic | File | Key Points |
|-------|------|------------|
| HuggingFace | [huggingface.md](huggingface.md) | Hub API, model download, caching, authentication |
## Module Structure
```
common/
├── pyproject.toml # UV-managed dependencies
├── llamafarm_common/
│ ├── __init__.py # Public API exports
│ ├── model_utils.py # GGUF file utilities
│ └── pidfile.py # PID file management
└── tests/
└── test_model_utils.py # Unit tests with mocking
```
## Public API
### Model Utilities
```python
from llamafarm_common import (
# Parse model:quantization syntax
parse_model_with_quantization,
# Extract quantization from filename
parse_quantization_from_filename,
# Select best GGUF file from list
select_gguf_file,
select_gguf_file_with_logging,
# List GGUF files in HF repo
list_gguf_files,
# Download and get path to GGUF file
get_gguf_file_path,
# Default quantization preference order
GGUF_QUANTIZATION_PREFERENCE_ORDER,
)
```
### PID File Management
```python
from llamafarm_common.pidfile import write_pid, get_pid_file
```
## Review Checklist Summary
When reviewing code in `common/`:
1. **HuggingFace Integration** (High priority)
- Proper error handling for network failures
- Authentication token passed correctly
- High-speed transfer enabled appropriately
2. **Model Selection** (Medium priority)
- Quantization preference order maintained
- Case-insensitive matching
- Graceful fallback when preferred not available
3. **Testing** (High priority)
- HuggingFace API calls mocked
- Network isolation in tests
- Edge cases covered (empty lists, missing files)
4. **Security** (Medium priority)
- No token exposure in logs
- Safe file path handling
- Environment variable protection
See [huggingface.md](huggingface.md) for detailed HuggingFace-specific checklists.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |