agentleFS
Sign inStart free
pooriaarabpooriaarab/skills/eco-mode/SKILL.md

eco-mode

Skill1 starsChanged 6 days agoMITView on GitHub

Reduce token/CO₂ usage without quality loss by compressing output, batching tool calls, and suggesting cheaper model tiers. Use when the user says 'eco mode' or runs /eco.

  • Installs packages

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    All levels preserve 100% technical accuracy (Jegham et al.

    line 22
  2. Distinctive:

    When eco mode is active:

    line 26
  3. Distinctive:

    Batch tool calls: read multiple files in one message, not sequentially.

    line 29
  4. Distinctive:

    Avoid re-reads: never read a file you've already read this session.

    line 30
  5. Distinctive:

    Before each tool call: ask "is this necessary?"

    line 43
  6. Distinctive:

    No summaries. No "I'll now...".

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Eco Mode

Activate: /eco or "eco mode" or "enable eco mode"

Deactivate: "stop eco mode" or "normal mode"

Based on Jegham et al. (arXiv:2505.09598, 2025): Claude Sonnet 4.6 emits 0.85g CO₂ per 1,000 tokens. Claude Haiku 4.5 emits 0.10g CO₂ per 1,000 tokens — 8.5× less.

Intensity Levels

LevelTriggerSavings
lite/eco lite~30% — drop filler, keep grammar
full/eco~65% — compress + batch tool calls + model suggestions
ultra/eco ultra~80% — maximum discipline, terse output

All levels preserve 100% technical accuracy (Jegham et al. 2025 / caveman benchmarks).

Eco Full (default)

When eco mode is active:

  • Compress output: drop filler, hedging, pleasantries. Fragments OK. 100% technical substance.
  • Batch tool calls: read multiple files in one message, not sequentially.
  • Avoid re-reads: never read a file you've already read this session.
  • Flag model downgrade opportunities: claude-haiku-4-5 uses 8.5× less CO₂ than claude-sonnet-4-6 (0.10g vs 0.85g per 1K tokens). Note when a sub-task could use a smaller model.

Pattern: [finding] [action] [reason]. No throat-clearing.

Eco Lite

Drop filler words, pleasantries, hedging. Keep all technical substance. No trailing summaries of what you just did.

Eco Ultra

  • Terse output only. Pattern: [thing] [action] [reason]. [next step]. Fragments. No filler.
  • Batch ALL tool calls. Never read a file twice.
  • Before each tool call: ask "is this necessary?"
  • Flag downgrade opportunities: Haiku 0.10g/1Ktok vs Sonnet 0.85g/1Ktok vs Opus 0.55g/1Ktok.
  • No summaries. No "I'll now...". No "Great, I've...". Just the output.

Why use many token when few do trick.

Also available as a vibenotifications plugin

npm install -g vibenotifications
vibenotifications add eco

Shows ♻️ ECO FULL · ~65% token savings in your Claude Code status line and injects eco mode instructions into every session automatically.

Source: github.com/itsryanarmani/vibenotifications

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free