Stxyu-p

Trimtok — AI skill for Claude Code

AI community

Universal, deterministic context compressor for AI coding agents & IDEs (Claude Code, Cursor, Antigravity, Codex, Gemini, Hermes) — 88.3% token reduction, lossless TOON v2 JSON tables, AST code skelet.

How to install Trimtok

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Stxyu-p/trimtok and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Trimtok does

Universal, deterministic context compressor for AI coding agents & IDEs (Claude Code, Cursor, Antigravity, Codex, Gemini, Hermes) — 88.3% token reduction, lossless TOON v2 JSON tables, AST code skeletonization, and SQLite WAL + Zstd CCR (#c1) recovery over stdio MCP.

Alternatives in AI

  • Jcodemunch MCP — Cut AI token costs 95%+ on code exploration 2.6k ★
  • Claude Fable 5 System Prompt Clean — the optimized, token-efficient version of the leaked Claude Fable 5 / Mythos 5 system prompt 475 ★
  • Engrim — The Universal Cross-Model Episodic Memory Standard 219 ★

README

trimtok

**Universal, Local, Deterministic Context Compressor for AI Coding Agents & IDEs**

88.3% fewer characters across heavy engineering workloads — zero intelligence lost, every byte recoverable in sub-millisecond time.

PyPI Latency 100% Local & Offline Fail-open MCP · CLI · Hermes License

Why trimtok · Before / After · Quick Start · How It Works · CCR Recovery · v4.2 Engine · Architecture · Performance · Configuration · Trinity Combo


Why trimtok?

Autonomous coding agents burn 60%–85% of their context window on **high-volume, low-entropy noise**: 2,000-line passing `pytest` runs, deep framework stack frames (`site-packages`, `node_modules`), unchanged `git diff` context lines, repetitive `grep` matches, and deeply nested JSON tool responses.

Naive truncation (`head -n 50` or character slicing) destroys critical failure signals hidden in the middle of logs. Summarizing via a secondary LLM adds seconds o