#123123rd of 389 in Skills, plugins and prompts
tokeneconomizer
Plugin that cuts coding-agent token cost using per-repository learned experience
- Stars
- 5.7k
- License
- Apache-2.0
- Last commit
- Nov 2026
- Last release
- Oct 2026
- Language
- Python
50 on 13 Oct 20265,708 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
TokenEconomizer is a plugin for Claude Code, Codex, GitHub Copilot CLI and Pi, with experimental support for several other agents. It learns from past sessions in a repository and applies three mechanisms: micro-skills (repeated facts injected at session start), tool pruning (unused tools left out via agent settings) and a loop guard that redirects stuck turns. The authors report 18.4–30.9% lower mean cost on SWE-Bench tasks across five agent–model combinations.
Who it is for: Developers paying for Claude Code, Codex or Copilot CLI in repeated repo work
Strengths
- Works through agent settings and context, so the agent itself is unchanged
- Reported 18.4–30.9% lower mean cost on SWE-Bench; its own model calls are counted
- Needs no API key of its own; uses your agent's existing login
- Actions are reversible; learned data stays local in ~/.tokeneconomizer
Weaknesses
- Micro-skills is off by default; savings depend on the repository and work
- Pinned to narrow tested version ranges; newer agent versions are untested
- Experimental agents lack features; Grok, Hermes and Antigravity have no tool pruning
- Cursor and Antigravity support is macOS and Linux only; Cursor reports no session cost
What it needs
- no GPU
- Needs Python 3.10+, A supported coding agent (Claude Code, Codex, Copilot CLI or Pi)
- Models: Claude Code, Codex, GitHub Copilot CLI, Pi, opencode (experimental)
Quick answers
Is tokeneconomizer open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Also in Skills, plugins and prompts
See all 389| Rank | Project | Score |
|---|---|---|
| 1 | claude-memPersistent cross-session memory plugin for Claude Code and other coding agents | 97 out of 100 |
| 2 | impeccableDesign skill with 24 slash commands for AI coding agents' frontend output | 97 out of 100 |
| 3 | gstackMarkdown slash-command skills that give Claude Code planning, review, QA and ship roles | 96 out of 100 |
| #4 | rufloAgent harness adding swarms, memory and hooks to Claude Code and Codex | 96 out of 100 |
| #5 | cavemanShortens coding-agent replies and compresses tool output before the model reads it | 95 out of 100 |