skills.rest
ttphats
ttphatsCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

token-budget-advisor

Estimates prompt token counts and offers selectable response depth levels before answering.

Updated Jan 28, 2026
Search Tags:#token budget#cost control#token estimation#response depth#prompt complexity#answer length
One-click install
npx skills add https://github.com/ttphats/project-detedxs26 --skill token-budget-advisor-ttphats
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: token-budget-advisor
Source: https://github.com/ttphats/project-detedxs26/tree/main/.augment/skills/ecc/token-budget-advisor
Command: npx skills add https://github.com/ttphats/project-detedxs26 --skill token-budget-advisor-ttphats

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Users often receive answers that are too long or too short for their needs, wasting tokens and time. This Skill intercepts the response flow to let users choose how much depth and length an answer should consume before it is generated. ## Core Features & Use Cases - Heuristic Token Estimation: Estimates input tokens using word-count and character-count heuristics, then projects a response window based on prompt complexity. - Four Depth Levels: Offers Essential (25%), Moderate (50%), Detailed (75%), and Exhaustive (100%) response tiers with estimated token counts for each. - Shortcut Detection: Recognizes direct signals like "short version", "tldr", or "50% depth" and responds at that level immediately without asking. - Use Case: A user asks a complex architecture question but only wants a brief summary first. The Skill presents depth options with estimated token costs, the user picks 25%, and receives a concise direct answer. ## Quick Start Ask the assistant to give you the short version of its answer or specify a depth level like 50% before it responds.

Frequently Asked Questions about token-budget-advisor

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

FAQPage Schema
How do I control the length of an AI response?▼

State a depth level directly, such as "short version", "tldr", or "50% depth", and the response is generated at that level immediately. Alternatively, the Skill presents four options from Essential (25%) to Exhaustive (100%) with estimated token counts before answering.

How are token counts estimated without a tokenizer?▼

The Skill uses heuristic rules: prose is estimated at words multiplied by 1.3, and code-heavy content at characters divided by 4. Accuracy is approximately 85-90% with a variance of plus or minus 15%, and a disclaimer is always shown.

What response depth levels are available?▼

Four levels exist: Essential at 25% gives a direct 2-4 sentence answer, Moderate at 50% adds context and one example, Detailed at 75% includes alternatives and pros/cons, and Exhaustive at 100% provides unrestricted full analysis.

Does the chosen depth level persist across a session?▼

Yes. Once a user selects a level, it is maintained silently for all subsequent responses in that session unless the user explicitly changes it. The Skill does not re-ask the depth question after a level has been set.

When does the token budget advisor not trigger?▼

It does not trigger when the user already set a level in the current session, when the answer is trivially one line, or when "token" refers to authentication, session, or payment tokens rather than response size.

Related Skills

Explore More Personal & Entrepreneur Skills→
MaelwalserMaelwalser

token-budget-advisor

Estimate input tokens and present four depth levels with token estimates.

Community
Intermediate
cescraflicescrafli

token-budget-advisor

Estimate input tokens and present response depth options before answering.

Community
Advanced
caovinhphuccaovinhphuc

token-budget-advisor

Intercept response flow to present token-based depth options before replying.

Community
Advanced
GGEduGGEdu

token-budget-advisor

Prompt users to select a depth level before generating responses.

Community
Advanced