Open-Source & Systems Projects
Autonomous SRE telemetry, Model Context Protocol (MCP) tools, and security-first engineering platforms
DAG-Ops: Deterministic AI Orchestration Engine
A dependency-free Python distributed DAG execution engine applying workflow graph principles and Kahn's topological sort (1962) to enterprise LLM pipelines, eliminating monolithic prompt fragility.
Key Highlights:
• Compile-time cycle detection and execution sequencing via Kahn's topological sort (1962).
• Concurrent parallel branch execution with ThreadPoolExecutor, cutting latency by 42%.
• Localized fault boundaries: isolated retry policy on transient HTTP 504 errors with zero blast radius.
• Pre-model deterministic Python regex firewall protecting PII and database secrets.
Git Systems Architecture & 3-Way Merge Conflict Engine
An interactive systems engineering platform and visual state simulator dissecting Git's content-addressable Merkle DAG, 4-zone state machine, and 3-way Lowest Common Ancestor (LCA) merge algorithm.
Key Highlights:
• Interactive 4-zone state machine: Working Tree → Index/Staging (.git/index) → Local Repo (.git/objects) → Remote.
• Live 3-way merge conflict editor simulating Base (LCA), Ours, and Theirs diffing with automatic and manual resolution paths.
• Deep-dive systems architecture modules on zlib compression, 41-byte branch pointers, and git reflog disaster recovery.
• Enterprise merge conflict governance models reducing trunk-based release integration bottlenecks by 45%.
ForgeOps AI
A safety-first autonomous reliability platform designed to act as an unprivileged, read-only SRE agent over computing infrastructure via Model Context Protocol (MCP).
Key Highlights:
• Deterministic policy engine preventing confused deputy execution.
• Automatic sanitization of diagnostic logs to neutralize indirect prompt injection attacks.
• Secret masking daemon stripping JWTs, AWS keys, and database URIs before telemetry ingestion.
FastMCP Linux Observability Server
A lightweight Model Context Protocol microservice exposing realtime CPU, RAM, and storage health metrics over Server-Sent Events (SSE) and local stdio, enforcing zero-root diagnostic access.
Key Highlights:
• Zero-dependency collector for Linux host metrics (load average, memory breakdown, active socket mappings).
• Native SSE and stdio transport support for seamless integration with LLMs and AI agent runtimes.
Enterprise Mail & Incident Intelligence Pipeline
Automated incident triage and ticket ingestion engine for enterprise Linux and mail environments, combining email security authentication with local LLM routing.
Key Highlights:
• Extracts and validates SPF, DKIM, and DMARC metadata to detect email spoofing.
• Intelligent privacy routing: local on-premise Ollama instances handle confidential PII, while frontier models handle escalations, cutting API costs by 85%.
Python Systems Engineering & CLI Toolkit
Architectural guides and modular CLI tools demonstrating production-grade Python design patterns for infrastructure automation and systems administration.
Key Highlights:
• Deep dives on Linux file systems, inode mechanics, and path traversals.
• Clean CLI design with argparse and standard library utilities without bloated runtime dependencies.