Sdnd Dev

C 62 completed
Framework
unknown / python · tiny
15
Files
666
LOC
0
Frameworks
2
Languages

Pipeline State

completed
Run ID
#370447
Phase
done
Progress
1%
Started
Finished
2026-04-13 01:31:02
LLM tokens
0

Pipeline Metadata

Stage
Cataloged
Decision
proceed
Novelty
40.55
Framework unique
Isolation
Last stage change
2026-05-10 03:35:28
Deduplication group #51189
Member of a group with 16 similar repo(s) — canonical #25995 view group →
Top concepts (1)
Project Description
About: code-quality intelligence by Repobility · https://repobility.com

AI Prompt

Create a framework in Python that allows multiple AI agents to autonomously generate, execute, and review code in a collaborative cycle. I need the core functionality to simulate three agents—Architect, Implementer, and Reviewer—interacting to improve software. The system should have a main development session entry point, a code execution sandbox, and modules for running benchmarks, including a self-improvement cycle (RSI). Please structure the project using the provided directory layout, ensuring it can interact with an external LLM backend, like one provided by `ollama_client.py`.
python ai-agents framework autonomy code-generation llm software-development benchmarking
Generated by gemma4:latest

Catalog Information

This framework enables multiple AI agents to autonomously generate, execute, and review code in a collaborative cycle, streamlining software development.

Description

The framework orchestrates a trio of AI agents—Architect, Implementer, and Reviewer—to autonomously produce, run, and critique code. It leverages a local large‑language‑model backend to generate code snippets, execute them in a sandboxed environment, and evaluate the results. Developers interact with the system through a command‑line interface, specifying tasks and reviewing the iterative improvements. The cycle continues until the code meets predefined quality criteria or the user stops the session. This approach reduces manual debugging, accelerates prototyping, and provides a structured pathway for AI‑driven software creation.

الوصف

يُنسق الإطار ثلاث وكلاء ذكاء اصطناعي—المهندس المعماري، والمُنفذ، والمراجع—لتوليد الكود وتشغيله وتقييمه بشكل ذاتي. يعتمد على بيئة محلية للذكاء الاصطناعي لتوليد مقاطع الكود، ثم يُنفّذها داخل حاوية آمنة، ويُقيّم النتائج. يتفاعل المطورون مع النظام عبر واجهة سطر أوامر، حيث يُحدّدون المهام ويستعرضون التحسينات المتتالية. تستمر الدورة حتى يحقق الكود معايير الجودة المحددة أو يتوقف المستخدم. يساهم هذا الأسلوب في تقليل الأخطاء اليدوية، وتسريع عملية النمذجة الأولية، وتوفير مسار منظم لتطوير البرمجيات المدعومة بالذكاء الاصطناعي. يتيح الإطار أيضاً تسجيل جلسات التطوير والنتائج لتسهيل التحليل والمراجعة. يختلف عن الحلول التقليدية بتركيزه على التعاون بين الوكلاء وتكرار العملية تلقائياً دون تدخل بشري مستمر.

Novelty

8/10

Tags

ai-agent-collaboration autonomous-code-generation software-self‑improvement llm‑driven-development automated-code-review interactive-development-sessions sandboxed-execution

Claude Models

claude-opus-4.6

Quality Score

C
62.5/100
Structure
44
Code Quality
98
Documentation
60
Testing
0
Practices
68
Security
100
Dependencies
60

Strengths

  • Consistent naming conventions (snake_case)
  • Good security practices \u2014 no major issues detected

Weaknesses

  • No LICENSE file \u2014 legal ambiguity for contributors
  • No tests found \u2014 high risk of regressions
  • No CI/CD configuration \u2014 manual testing and deployment

Recommendations

  • Add a test suite \u2014 start with critical path integration tests
  • Set up CI/CD (GitHub Actions recommended) to automate testing and deployment
  • Add a linter configuration to enforce code style consistency
  • Add a LICENSE file (MIT recommended for open source)

Security & Health

4.6h
Tech Debt (E)
A
OWASP (100%)
PASS
Quality Gate
A
Risk (10)
All rows scored by the Repobility analyzer (https://repobility.com)
Unknown
License
3.2%
Duplication
Full Security Report AI Fix Prompts SARIF SBOM

Languages

python
92.6%
markdown
7.4%

Frameworks

None detected

Concepts (1)

Per-row analysis by Repobility · https://repobility.com
CategoryNameDescriptionConfidence
Repobility · code-quality intelligence · https://repobility.com
auto_descriptionProject Description複数AIエージェントが協調してソフトウェアを自律改善する実験的フレームワーク。 sdnd-theater(TRPGセッション生成)の姉妹プロジェクト。80%

Quality Timeline

1 quality score recorded.

View File Metrics

Embed Badge

Add to your README:

![Quality](https://repos.aljefra.com/badge/94652.svg)
Quality BadgeSecurity Badge
Export Quality CSVDownload SBOMExport Findings CSV