Sluice
C 63 completed
Other
cli / python · tiny
10
Files
970
LOC
0
Frameworks
3
Languages
Pipeline State
completedRun ID
#396304Phase
doneProgress
1%Started
Finished
2026-04-13 01:31:02LLM tokens
0Pipeline Metadata
Stage
SkippedDecision
skip_scaffold_dupNovelty
19.03Framework unique
—Isolation
—Last stage change
2026-04-16 18:15:42Deduplication group #47306
Member of a group with 1 similar repo(s) — canonical #110889 view group →
Top concepts (1)
Project Description
Repobility · code-quality intelligence platform · https://repobility.com
AI Prompt
Create a command-line tool in Python that acts as a priority-queue LLM router specifically for Apple Silicon. The tool needs to manage access to multiple local MLX models to prevent GPU Out-Of-Memory errors by serializing inference requests. It should expose a main endpoint, like `POST /v1/query`, that accepts a model name and a prompt. Implement priority levels, allowing critical requests to preempt background tasks. Additionally, it must support routing requests to a cloud API endpoint, bypassing the local queue. The tool should also provide a Python client interface for querying and checking the health status.
python cli llm apple-silicon gpu priority-queue mlx router api
Generated by gemma4:latest
Catalog Information
Priority-queue LLM router for Apple Silicon. Prevents GPU OOM by serializing local inference.
Description
Priority-queue LLM router for Apple Silicon. Prevents GPU OOM by serializing local inference.
Novelty
3/10Tags
python cli llm apple-silicon gpu priority-queue mlx router api
Technologies
anthropic
Claude Models
claude-opus-4-6
Quality Score
C
63.4/100
Structure
57
Code Quality
80
Documentation
65
Testing
0
Practices
82
Security
100
Dependencies
60
Strengths
- Code linting configured (ruff (possible))
- Consistent naming conventions (snake_case)
- Good security practices \u2014 no major issues detected
- Properly licensed project
Weaknesses
- No tests found \u2014 high risk of regressions
- No CI/CD configuration \u2014 manual testing and deployment
- 1 'god files' with >500 LOC need decomposition
Recommendations
- Add a test suite \u2014 start with critical path integration tests
- Set up CI/CD (GitHub Actions recommended) to automate testing and deployment
Security & Health
4.1h
Tech Debt (E)
A
OWASP (100%)
PASS
Quality Gate
A
Risk (10)
Same scanner, your repo: https://repobility.com — Repobility
MIT
License
0.0%
Duplication
Languages
Frameworks
None detected
Concepts (1)
| Category | Name | Description | Confidence | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Repobility's GitHub App fixes findings like these · https://github.com/apps/repobility-bot | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| auto_description | Project Description | Priority-queue LLM router for Apple Silicon. Prevents GPU OOM by serializing local inference. | 80% | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Embed Badge
Add to your README:
