skills.rest
Dubibubii
DubibubiiCOMMUNITY
Explore all skills by this author →
GitHub

usage-limit-reducer

Analyze local JSONL logs to measure token usage by model, project, and day.

186|37|Updated Apr 19, 2026
Search Tags:#optimization#diagnostics#claude-code#cost#token-usage#usage-report
One-click install
npx skills add https://github.com/Dubibubii/usage-limit-reducer --skill usage-limit-reducer-dubibubii
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: usage-limit-reducer
Source: https://github.com/Dubibubii/usage-limit-reducer/tree/main
Command: npx skills add https://github.com/Dubibubii/usage-limit-reducer --skill usage-limit-reducer-dubibubii

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Claude Code re-reads the entire conversation on every turn, causing tokens to be wasted on history. This skill analyzes token usage and applies practical rules to cut unnecessary usage, improving efficiency and responsiveness.

Core Features & Use Cases

  • Measure: runs the bundled usage-report.py against local logs to show token/cost breakdowns by model, project, and day.
  • Diagnose & Apply: assesses session length, CLAUDE.md presence, and active model, then applies the most relevant rules to reduce waste.
  • Actionable Guidance: suggests concrete next steps (e.g., /compact, /clear, /model, create CLAUDE.md) and acts only with user confirmation.
  • Standalone Utility: the token report can run independently to surface per-day usage and costs, strictly offline.

Quick Start

Run the bundled usage-report to surface token usage and guide actionable reductions.

Frequently Asked Questions about usage-limit-reducer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reduce Claude Code token usage from re-reading conversation history?▼

To reduce Claude Code token usage from re-reading history, this skill analyzes local JSONL logs to measure per-model and per-day waste, then applies 11 usage-reduction rules to suggest actions like /compact or /clear for your current session context.

What is the best way to analyze Claude Code token costs per project?▼

The best way to analyze Claude Code token costs per project is by running the standalone usage-report.py script against local JSONL logs, which processes data strictly offline to surface breakdowns by model, project, and day.

Does the usage-limit-reducer skill make external network calls to fetch my token reports?▼

No, the usage-limit-reducer skill does not make external network calls; it operates strictly offline using local-only data from JSONL logs to generate token usage reports and tailored reduction guidance without sending data externally.

How do I get actionable guidance to cut wasted tokens in my active Claude Code session?▼

You get actionable guidance to cut wasted tokens by having the skill assess your session length, CLAUDE.md presence, and active model, then applying rules to suggest concrete steps like creating a CLAUDE.md or switching models, acting only with confirmation.

When should I use /compact or /clear commands to optimize Claude Code costs?▼

You should use /compact or /clear to optimize Claude Code costs when the skill diagnoses excessive token waste from lengthy conversation history, providing a concise plan with expected outcomes based on your specific session context and active model.

Related Skills

Explore More Software Engineering Skills→
dakshrawat298-gifdakshrawat298-gif

session-logs

Analyze session JSONL logs with jq and rg to extract costs, token usage, and tool interactions.

Community
Intermediate
majedsiefalnasrmajedsiefalnasr

rtk-execution-layer

Prefix shell commands with RTK to produce structured, token-optimized outputs.

Community
Intermediate
malinskibeniaminmalinskibeniamin
3

setup-agent-config

Configure Claude Code session hooks to minimize prompt length and context bloat.

Community
Intermediate
kubonykubony
10

gcp-usage

Reports and analyzes GCP API usage and costs from BigQuery billing_export data.

Community
Intermediate