Back to Home
m1rl0k icon

Context-Engine

Verified Safe

by m1rl0k

Overview

A Retrieval-Augmented Generation (RAG) stack for codebases, enabling context-aware AI agents for developers and IDEs through unified code indexing, hybrid search, and local LLM integration.

Installation

Run Command
docker compose up -d

Environment Variables

  • QDRANT_URL
  • COLLECTION_NAME
  • EMBEDDING_MODEL
  • LLAMACPP_URL
  • GLM_API_KEY
  • QDRANT_API_KEY
  • REFRAG_DECODER
  • RERANKER_ONNX_PATH
  • RERANKER_TOKENIZER_PATH
  • FASTMCP_HOST
  • FASTMCP_INDEXER_PORT
  • FASTMCP_PORT
  • USE_GPU_DECODER
  • MICRO_BUDGET_TOKENS

Security Notes

The project uses `subprocess` to execute Python scripts, which is safe if commands and arguments are controlled. Input validation and path sanitization (e.g., preventing directory traversal in `qdrant_index` and `upload_service`) are critical for security. The `upload_service` handles incoming tar archives, requiring robust validation to prevent malicious file uploads. `_ast.literal_eval` is used for lenient argument parsing but is generally safer than `eval`. Default CORS settings (`*`) for the upload service and `HostPath` volumes in Kubernetes deployments require careful consideration in production environments.

Similar Servers

Stats

Interest Score82
Security Score7
Cost ClassLow
Avg Tokens750
Stars92
Forks9
Last Update2025-12-05

Tags

RAGCode IndexingVector SearchLLM IntegrationIDE ToolingDeveloper ExperienceKubernetesMicroservicesSemantic Search