https://github.com/profrandom92/comptext-cli
https://github.com/profrandom92/comptext-cli
Last synced: 26 days ago
JSON representation
- Host: GitHub
- URL: https://github.com/profrandom92/comptext-cli
- Owner: ProfRandom92
- Created: 2026-05-06T07:35:35.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2026-05-07T09:00:29.000Z (about 2 months ago)
- Last Synced: 2026-05-07T09:31:41.924Z (about 2 months ago)
- Language: Python
- Size: 88.9 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# CompText Safe CI + MCP Test Harness
## Overview
CompText is a professional, self-learning adaptive AI kernel. This repository implements a **Safe CI Layer** and **MCP Test Harness** to ensure the evolution of the kernel remains deterministic and stable during integration.
- **Deterministic CI Layer**: Prevents system drift during automated testing.
- **MCP Mocked Environment**: Fully simulates Context7 and Render MCPs.
- **Zero-Drift Validation**: Enforces safety bounds on system learning.
## Architecture
The CompText system is composed of several intelligent layers working in concert:
- **Kernel**: The core self-learning system implementing adaptive reasoning.
- **Router**: Adaptive scoring engine that maps intents to actions.
- **Agent Loop**: The main execution pipeline coordinating kernel logic.
- **Tool Intelligence**: Performance-aware selection of internal and external tools.
- **MCP Bridge**: Unified interface for Context7 (Semantic) and Render (Execution) MCPs.
- **Stability Guard**: Stability layer that monitors system drift and enforces operational bounds.
- **Safe CI Harness**: The testing infrastructure implemented in this repository.
### System Flow
```
CompText Kernel
↓
Router (Adaptive Scoring)
↓
Agent Loop
↓
Tool Intelligence → MCP Bridge
↓
Context7 / Render (Mocked in CI)
```
## Safe CI Mode
During CI/CD execution, the system operates in a specialized **Safe Mode**:
- **No Learning**: Router weights and Brain performance metrics are not updated.
- **MCP Mocks**: All external MCP calls are intercepted and handled by the Mock Layer.
- **Deterministic Execution**: Randomized behaviors are disabled to ensure reproducibility.
- **Isolation**: Test runs do not persist changes to the production memory state.
## MCP Mock System
The harness provides specialized mocks for external dependencies:
- **Context7 Mock**: Provides stable, semantic retrieval results for validation.
- **Render Mock**: Simulates execution runtimes and deployment feedback with guaranteed latency profiles.
- **Reproducibility**: Guarantees that the same input always produces the same output in the CI environment.
## Testing Strategy
- **Unit Tests**: Direct validation of Router logic, Memory constraints, and DAG execution.
- **CI Regression Tests**: Automated verification of core kernel behaviors on every push.
- **Drift Prevention**: Ensures that architectural changes do not violate the stability bounds.
- **Deterministic DAG Checks**: Validates dependency resolution and parallel execution patterns.
## Why This Matters
Adaptive AI systems are prone to **system drift**—where learning from a specific environment degrades performance in another. This CI infrastructure:
- Prevents runaway learning side effects.
- Ensures absolute reproducibility of system failures.
- Enables safe, rapid evolution of the kernel without risking production stability.
## CI Pipeline
The GitHub Actions workflow follows a strict validation path:
1. **Checkout**: Source code retrieval.
2. **Setup**: Python 3.11 environment initialization.
3. **Install**: Dependency resolution including `pytest` and `pytest-asyncio`.
4. **Pytest**: Execution of the deterministic CI test suite.
5. **Validate**: Final check of system stability.
## Safety Guarantee
> "The CI environment is fully isolated from the adaptive learning system and enforces deterministic execution across all components."
---
*CompText Kernel v1.0.0-ci*