terramentis-ai/terraqore-studio
Meta-Agentic Ai Orchestration Platform For Developers and Data Scientists Developing Full-Cycle(E2E) AI Projects
What's novel
Meta-Agentic Ai Orchestration Platform For Developers and Data Scientists Developing Full-Cycle(E2E) AI Projects
Code Analysis
7 files read · 3 roundsA CLI-based multi-agent orchestration platform that chains LLM calls (via OpenRouter/Gemini/xAI/Ollama) through a fixed pipeline of prompt-engineered agents (ideation → planning → code generation → validation) to produce project plans and code files, with SQLite state tracking and a refinement loop.
Strengths
Well-structured agent base class with prompt profiles, security validation hooks, and RAG integration; the executor's iterative generate-validate-refine loop with a ConflictResolverAgent is a reasonable pattern; multi-provider LLM abstraction with fallback is practical.
Weaknesses
The 'agents' are almost entirely prompt templates + a single LLM call — there is no real algorithmic logic, no actual code analysis, no genuine conflict resolution; the PSMP system and many 'enterprise' features (secure gateway, compliance policies, achievement system) appear to be thin wrappers or aspirational scaffolding rather than working implementations; the codebase is bloated with 200+ source files for what is essentially a prompt-chaining CLI tool.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
1857
Contributors
2
Files
155
Active weeks
12
Repository
Language
Python
Stars
3
Forks
0
License
—