HoangP8/tokless

An effective and efficient coding-agent setup in under 30 seconds: best tools, unified instructions, any OS.

What it solves

Coding harnesses—tools that help AI agents write and manage code—often require complex, manual configuration to work effectively. Tokless solves this by acting as a unified pipeline that automatically wires together specialized, high-performance coding tools without requiring manual setup or conflicting configurations.

How it works

Tokless installs a suite of eight specialized packages and integrates them into your chosen coding harness (like Claude Code, Cursor, or Cline). It handles the technical "wiring" automatically: it injects instruction rules, configures MCP (Model Context Protocol) servers, sets up command hooks, and routes API traffic through a local proxy. It intelligently filters tool lists to remove unnecessary diagnostics, ensuring only essential tools land in your AI's context.

Who it’s for

This tool is for developers who use AI coding agents and want to optimize their performance, memory, and context management without spending time on manual configuration or troubleshooting tool conflicts.

Highlights

  • Zero-config installation that works in under 30 seconds.
  • Unified management for 16 different coding harnesses.
  • Automatically compresses command output and prompts to save context space.
  • Injects persistent project memory (decisions, fixes, and notes) into every session.
  • Provides indexed code graphs for better source and call-path analysis.
  • Automatically prunes redundant tools to keep AI context clean and focused.

Related

  • Project
    djcopley/ShellOracle

    A terminal utility that generates shell commands from natural language descriptions and inserts them directly into the prompt using various LLM providers.

  • Project
    zhu-guli326/image2_UI

    An open-source UI generation toolkit for OpenAI Codex that converts screenshots, design references, or text briefs into editable, interactive frontend prototypes.

  • Project
    SakuraByteCore/codexmate

    Codex Mate is a local-first control plane that manages multiple AI coding agents — Codex, Claude Code, Gemini CLI, and others — from one CLI and web dashboard. It unifies session browsing, provider switching, prompt editing, and task orchestration while keeping all data on your machine.

  • Project
    clay-good/OpenLore

    OpenLore builds a deterministic, local-first knowledge graph of your codebase so AI coding agents start every task oriented and get their changes certified before landing — no API key, no LLM in the hot path.