Skip to content
View aloneangel45's full-sized avatar
1 follower · 2 following

Achievements

Achievement: Pull SharkAchievement: YOLO

Achievements

Achievement: Pull SharkAchievement: YOLO

Block or report aloneangel45

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
aloneangel45/README.md

Hi there, I'm Arda (@aloneangel45)

Building reliable, deterministic infrastructure for AI agents and developer tooling.
Engineering high-throughput, deterministic runtime proxies, checkpointing engines, and low-level system utilities for autonomous AI development workflows.

ToolPrune v1.2.0 TracePort v1.1.0 AgentFixture v1.1.0 FocusFlow v2.1.0

Featured Ecosystem: Agentic Infrastructure & Systems Tooling

A tightly integrated, zero-bloat developer toolchain engineered to compress context-window overhead, guarantee deterministic replay, and protect deep-work focus:

Project Role & Purpose Key Highlights CI Stack
ToolPrune Adaptive MCP Proxy & Token-Budget Stream Filter Progressive --max-schema-tokens compaction and dynamic tool hydration cutting MCP schema tokens by 93.3%–96.5% at <0.4ms cold-pass latency across Stdio, SSE, and Streamable HTTP (/mcp). CI Go MCP JSON-RPC
TracePort Deterministic Flight Recorder & Zstd Shadow-Tree Engine Content-addressed workspace checkpointing with Zstd preset/trained dictionaries (86.1% footprint reduction, 557us stat-cached scan) and O(1) JSONL flight logs for $0 zero-token rollback. CI Go Zstd Git
AgentFixture Zero-Cost Deterministic CI/CD Cassette Runner Native Anthropic & OpenAI SSE stream recording/replay (instant/scaled/exact timing), zero-outbound-network guard, 16MB MCP stdio proxy, Go agenttest SDK, and TracePort v1.1 importer. CI Go SSE CI/CD
FocusFlow Deep Work Discipline Engine & Desktop HUD Unified Go CLI and zero-dependency Python Catppuccin Mocha HUD with crash-safe SHA-256 dual-stack IPv4/IPv6 hosts sinkholing, Win32 DACL preservation, and low-overhead Win32/POSIX process supervision. CI Go Python Win32

Ecosystem Architecture Workflow

   ┌───────────────────────────────────────────────────────────────┐
   │        AI Coding Agent (Claude Code / Cursor / Windsurf)      │
   └───────────────────────────────┬───────────────────────────────┘
                                   │
          ┌────────────────────────┴────────────────────────┐
          ▼                                                 ▼
┌─────────────────────┐                           ┌─────────────────────┐
│      ToolPrune      │                           │      TracePort      │
│  (Adaptive Stream   │                           │  (Zstd Shadow-Tree  │
│  93.3%-96.5% Prune) │                           │  & Flight Recorder) │
└──────────┬──────────┘                           └──────────┬──────────┘
           │                                                 │
           ▼                                                 ▼
 [Upstream MCP Fleet]                             ┌─────────────────────┐
 (Stdio / SSE / /mcp)                             │    AgentFixture     │
                                                  │ (SSE + MCP Offline  │
                                                  │   $0 CI Cassettes)  │
                                                  └─────────────────────┘

Core Tech Stack & Engineering Focus

  • Languages: Go, Rust, Python, C/C++, TypeScript, Modern Shell / PowerShell
  • Systems & Low-Level: Win32 API (Toolhelp32, ReplaceFileW DACLs, Job Objects), POSIX (/proc, process groups, signal handling), IPC (Stdio, SSE, Streamable HTTP)
  • Agentic & Protocols: Model Context Protocol (MCP), JSON-RPC 2.0, Anthropic & OpenAI Server-Sent Events (SSE), cl100k_base Token Calibration, Pre-Bind JSON Schema Validation
  • Storage & Determinism: Dictionary-Trained Zstandard (Zstd) Content-Addressable Storage, Append-Only JSONL Event Logs, Crash-Safe Write-Ahead Journals, SHA-256 Integrity Verification
  • DevOps & QA: Multi-OS GitHub Actions Matrices (Linux, Windows, macOS), Go Race Detector Auditing, Reproducible CGO_ENABLED=0 Cross-Compilation (amd64 / arm64)

GitHub Activity & Metrics

GitHub Stats GitHub Streak

"Build with zero bloat, verify with zero mocks, engineer for deterministic resilience."
Crafted by aloneangel45

Pinned Loading

  1. FocusFlow FocusFlow Public

    Cross-platform deep work discipline engine: atomic SHA-256 dual-stack IPv4/IPv6 hosts sinkholing, zero-dependency Win32/Linux/macOS process supervisor & Catppuccin Mocha HUD.

    Python

  2. AgentFixture AgentFixture Public

    Zero-dollar deterministic CI/CD cassette recorder, Anthropic/OpenAI SSE stream replayer, MCP stdio interceptor & Go testing SDK for AI agents.

    Go

  3. toolprune toolprune Public

    High-performance MCP proxy, adaptive token-budget stream filter & multi-upstream gateway. Cuts LLM tool schema tokens by up to 96.5% for Claude Code, Cursor & Windsurf.

    Go 1

  4. traceport traceport Public

    Deterministic flight recorder & Zstd dictionary shadow-tree checkpointing engine for autonomous AI coding agents (86% compressed footprint reduction).

    Go