New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Nowait Reasoning Optimizer

by davila7 in davila7/claude-code-templates

Implements the NOWAIT technique for efficient reasoning in R1-style LLMs.

MITAuto-check passedAI & LLM Engineering

Install Nowait Reasoning Optimizer

skills CLI
$ npx skills add davila7/claude-code-templates --skill nowait-reasoning-optimizer -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install davila7/claude-code-templates nowait-reasoning-optimizer --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/davila7/claude-code-templates.git skills-src && mkdir -p .claude/skills && cp -r skills-src/cli-tool/components/skills/productivity/nowait .claude/skills/nowait-reasoning-optimizer && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
nowait-reasoning-optimizer
GitHub stars
32k
Used in
2 other repos
Token cost
~1.2k tokens
SKILL.md length
334 words
Files
3 (incl. scripts)
Skills in repo
478
Repo updated
First seen
Licence
MIT

At a glance

Implements the NOWAIT technique for efficient reasoning in R1-style LLMs.

  • Works in 2 steps: Basic Implementation → Keywords Suppressed
  • Optimizing inference of reasoning models (QwQ
  • SKILL.md covers Overview, When to Use, Supported Models and Quick Start, plus 6 more sections
  • Runs Python scripts from its folder

What it does

Nowait Reasoning Optimizer is an agent skill from davila7/claude-code-templates. Implements the NOWAIT technique for efficient reasoning in R1-style LLMs. Use when optimizing inference of reasoning models (QwQ, DeepSeek-R1, Phi4-Reasoning, Qwen3, Kimi-VL, QvQ), reducing chain-of-thought token usage by 27-51% while preserving accuracy. Triggers on "optimize reasoning", "reduce thinking tokens", "efficient inference", "suppress reflection tokens", or when working with verbose CoT outputs.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 4 other files, including scripts (for example `refrences/keywords.md` and `scripts/nowait_processor.py`).

It sits in AI & LLM Engineering. It works with Qwen, Kimi and DeepSeek. The repository describes itself as: CLI tool for configuring and monitoring Claude Code. The licence is MIT.

When your agent uses it

  • Optimizing inference of reasoning models (QwQ
  • Reducing chain-of-thought token usage by 27-51% while preserving accuracy
  • Optimize reasoning
  • Reduce thinking tokens

Example prompts

  • “optimize reasoning”
  • “reduce thinking tokens”
  • “efficient inference”
  • “/nowait-reasoning-optimizer”

Requirements

  • Python 3

Workflow steps

2 steps, taken from the step headings in SKILL.md.

  1. Basic Implementation
  2. Keywords Suppressed

What it can do on your machine

Read from SKILL.md and the folder at commit 46b4d8b. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 1 file in scripts/ (Python), which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Nowait Reasoning Optimizer loads about 1.2k tokens when it runs. Until then it costs about 109 tokens; SKILL.md has 334 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~109
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from davila7/claude-code-templates at commit 46b4d8b, republished under its MIT licence (© davila7). 334 words, ~1,217 tokens.

Download SKILL.mdSave it as .claude/skills/nowait-reasoning-optimizer/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.
name
nowait-reasoning-optimizer
description
Implements the NOWAIT technique for efficient reasoning in R1-style LLMs. Use when optimizing inference of reasoning models (QwQ, DeepSeek-R1, Phi4-Reasoning, Qwen3, Kimi-VL, QvQ), reducing chain-of-thought token usage by 27-51% while preserving accuracy. Triggers on "optimize reasoning", "reduce thinking tokens", "efficient inference", "suppress reflection tokens", or when working with verbose CoT outputs.

NOWAIT Reasoning Optimizer

Implements the NOWAIT technique from the paper "Wait, We Don't Need to 'Wait'! Removing Thinking Tokens Improves Reasoning Efficiency" (Wang et al., 2025).

Overview

NOWAIT is a training-free inference-time intervention that suppresses self-reflection tokens (e.g., "Wait", "Hmm", "Alternatively") during generation, reducing chain-of-thought (CoT) trajectory length by 27-51% without compromising model utility.

When to Use

  • Deploying R1-style reasoning models with limited compute
  • Reducing inference latency for production systems
  • Optimizing token costs for reasoning tasks
  • Working with verbose CoT outputs that need streamlining

Supported Models

Model SeriesTypeToken Reduction
QwQ-32BRL-based16-31%
Phi4-Reasoning-PlusRL-based23-28%
Qwen3-32BRL-based13-16%
Kimi-VL-A3BMultimodal40-60%
QvQ-72B-PreviewMultimodal20-30%

Important: NOWAIT works best with RL-based models. Distilled models (Qwen3-4B/8B/14B) show degraded performance when reflection tokens are suppressed.

Quick Start

1. Basic Implementation
python
from scripts.nowait_processor import NOWAITLogitProcessor

# Initialize processor for your model's tokenizer
processor = NOWAITLogitProcessor(tokenizer)

# Use during generation
outputs = model.generate(
    inputs,
    logits_processor=[processor],
    max_new_tokens=32768
)
2. Keywords Suppressed

See references/keywords.md for the complete list. Core keywords:

wait, alternatively, hmm, but, however, check, 
double-check, maybe, verify, again, oh, ah

How It Works

  1. Initialize Keywords: Identify reflection keywords from empirical analysis
  2. Expand to Token Variants: Map keywords to all token variants in vocabulary (e.g., "wait" → " wait", "Wait", " Wait", ".wait", "WAIT")
  3. Suppress During Inference: Set logits of reflection tokens to large negative values during decoding
Logits (Before)         Logits (After)
Wait     0.8     →     Wait     -inf
First    0.6     →     First    0.6
Hmm      0.5     →     Hmm      -inf
Let      0.4     →     Let      0.4

Key Findings

Why It Works
  • NOWAIT doesn't eliminate self-reflection entirely—it guides models to skip unnecessary "waiting" reasoning
  • Models still perform essential verification at key decision points
  • Results in more linear, straightforward reasoning paths
RL vs Distilled Models
Model TypeNOWAIT EffectRecommendation
RL-based (QwQ, Phi4, Qwen3-32B)Stable accuracy, significant token reduction✅ Recommended
Distilled (Qwen3-4B/8B/14B)Accuracy degradation on hard tasks⚠️ Use with caution

Distilled models rely heavily on CoT structure from training data—removing reflection tokens disrupts their reasoning patterns.

Integration Examples

HuggingFace Transformers
python
from transformers import AutoModelForCausalLM, AutoTokenizer
from scripts.nowait_processor import NOWAITLogitProcessor

model = AutoModelForCausalLM.from_pretrained("Qwen/QwQ-32B")
tokenizer = AutoTokenizer.from_pretrained("Qwen/QwQ-32B")

processor = NOWAITLogitProcessor(tokenizer)

response = model.generate(
    tokenizer(prompt, return_tensors="pt").input_ids,
    logits_processor=[processor],
    max_new_tokens=32768,
    do_sample=True,
    temperature=0.7
)
vLLM
python
from vllm import LLM, SamplingParams
from scripts.nowait_processor import get_nowait_bad_words_ids

llm = LLM(model="Qwen/QwQ-32B")
bad_words_ids = get_nowait_bad_words_ids(llm.get_tokenizer())

sampling_params = SamplingParams(
    max_tokens=32768,
    bad_words_ids=bad_words_ids
)

Expected Results

Task TypeOriginal TokensNOWAIT TokensReduction
Math (AIME)15,00010,50030%
Visual QA (MMMU)2,9001,45050%
Video QA (MMVU)1,7001,25027%

Limitations

  • Less effective on very simple problems where CoT overhead is already minimal
  • Distilled models may suffer accuracy loss on challenging tasks
  • Some domains may require model-specific keyword tuning

References

  • Paper: arXiv:2506.08343v2
  • Complete keyword list: references/keywords.md
  • Implementation: scripts/nowait_processor.py

© davila7, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 2 other files (scripts) in cli-tool/components/skills/productivity/nowait of davila7/claude-code-templates.

  • SKILL.md
  • refrences/keywords.md
  • scripts/nowait_processor.py

Open the folder on GitHubat commit 46b4d8b

Used in 2 other repositories

We found 2 copies of this SKILL.md (exact, near-identical or edited) in other folders, from 2 other GitHub owners. This page covers the copy in davila7/claude-code-templates, which our catalogue first saw on October 7, 2026.

Compare with similar skills

Nowait Reasoning Optimizer next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Nowait Reasoning Optimizer compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Nowait Reasoning Optimizer this skilldavila7/claude-code-templates32k2 repos~1.2kAutomated safety check: PassMIT
LLM Council on Fireworks AIdair-ai/dair-academy-plugins614—~5kAutomated safety check: NotesMIT
Claude Maintain ModelsKiln-AI/Kiln5.2k—~15kAutomated safety check: NotesCustom licence
LLM Pipeline Profiler AnalysisBBuf/AI-Infra-Auto-Driven-SKILLS925—~3.9kAutomated safety check: PassNone
Update Ollama Cloud Modelsheypinchy/pinchy182—~3.9kAutomated safety check: NotesAGPL-3.0
Model Routingalinaqi/maggy707—~1.5kAutomated safety check: PassMIT

Similar skills

  • LLM Council on Fireworks AI

    dair-ai/dair-academy-plugins

    Has several open-weight models answer a question, rank each other's anonymized answers, then lets a chairman model write the final response through Fireworks AI.

    614 GitHub stars~5k tokensUpdated 2 mo ago
    AI & LLM EngineeringAuto-check: notes

  • Claude Maintain Models

    Kiln-AI/Kiln

    Add new AI models to Kiln's mlmodellist.py and produce a Discord announcement.

    5.2k GitHub stars~15k tokensUpdated yesterday
    AI & LLM EngineeringAuto-check: notes

  • LLM Pipeline Profiler Analysis

    BBuf/AI-Infra-Auto-Driven-SKILLS

    Breaks LLM torch profiler traces down by forward pass, layer and kernel, with timing tables and Perfetto time ranges for the layers you want to inspect.

    925 GitHub stars~3.9k tokensUpdated 4 days ago
    AI & LLM EngineeringAuto-check passed

  • Update Ollama Cloud Models

    heypinchy/pinchy

    A skill your agent uses when a new Ollama Cloud model is announced or available (e.g.

    182 GitHub stars~3.9k tokensUpdated 18 days ago
    AI & LLM EngineeringAuto-check: notes

  • Model Routing

    alinaqi/maggy

    9-tier model routing system with cascading classifier fallback and result auto-evaluation

    707 GitHub stars~1.5k tokensUpdated 15 days ago
    AI & LLM EngineeringAuto-check passed

  • Vllm Daily PR Issue Tracker

    ascend-ai-coding/awesome-ascend-skills

    Track daily PRs and Issues from vllm-project/vllm and vllm-project/vllm-ascend, filter by model (DeepSeek/Qwen/GLM/MiniMax/Kimi) and tech topics (PD disaggregation, MTP, quantization, graph mode…

    174 GitHub stars~731 tokensUpdated today
    AI & LLM EngineeringAuto-check passed

More from davila7/claude-code-templates

All 478 skills in this repo

  • Perplexity Web Search

    davila7/claude-code-templates

    Runs web-grounded searches through Perplexity's Sonar models over OpenRouter for current events, recent literature and cited facts beyond the model's training cutoff.

    32k GitHub starsUsed in 11 repos~3.5k tokens
    Auto-check: notes

  • Neuropixels Data Analysis

    davila7/claude-code-templates

    Analyzes Neuropixels recordings from SpikeGLX or Open Ephys through preprocessing, drift correction, Kilosort4 spike sorting, quality metrics and curation.

    32k GitHub starsUsed in 9 repos~2.8k tokens
    Auto-check passed

  • Scientific Venue Templates

    davila7/claude-code-templates

    Supplies LaTeX templates and formatting rules for journals, conferences, posters, and grant proposals, then can check a draft against them.

    32k GitHub starsUsed in 9 repos~5.1k tokens
    Auto-check: notes

  • Brand Voice Content Creator

    davila7/claude-code-templates

    Analyzes a brand's existing writing to lock in a consistent voice, then builds SEO blog posts and platform-specific social content around it.

    32k GitHub starsUsed in 3 repos~1.9k tokens
    Auto-check passed

  • CAPA Officer

    davila7/claude-code-templates

    Guides corrective and preventive action (CAPA) work in a quality management system, from initiation and root cause analysis through effectiveness verification.

    32k GitHub starsUsed in 1 repo~2k tokens
    Auto-check passed

  • Fda Consultant Specialist

    davila7/claude-code-templates

    Senior FDA consultant and specialist for medical device companies including HIPAA compliance and requirement management.

    32k GitHub starsUsed in 1 repo~2.7k tokens
    Auto-check passed

Works with

Categories

Questions about Nowait Reasoning Optimizer

What does Nowait Reasoning Optimizer do?

Implements the NOWAIT technique for efficient reasoning in R1-style LLMs. Nowait Reasoning Optimizer is an agent skill from davila7/claude-code-templates. Implements the NOWAIT technique for efficient reasoning in R1-style LLMs.

When should I use Nowait Reasoning Optimizer?

Nowait Reasoning Optimizer fits situations like: optimizing inference of reasoning models (QwQ; reducing chain-of-thought token usage by 27-51% while preserving accuracy; optimize reasoning; reduce thinking tokens.

How do I install Nowait Reasoning Optimizer in Claude Code?

Run `npx skills add davila7/claude-code-templates --skill nowait-reasoning-optimizer -a claude-code`. Or copy the skill folder (cli-tool/components/skills/productivity/nowait in davila7/claude-code-templates) into .claude/skills/nowait-reasoning-optimizer in your project. Claude Code loads it when a task matches its description.

How do I install Nowait Reasoning Optimizer in Codex?

Run `npx skills add davila7/claude-code-templates --skill nowait-reasoning-optimizer -a codex`. Or copy the skill folder (cli-tool/components/skills/productivity/nowait in davila7/claude-code-templates) into .agents/skills/nowait-reasoning-optimizer in your project. Codex loads it when a task matches its description.

Can I use Nowait Reasoning Optimizer in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add davila7/claude-code-templates --skill nowait-reasoning-optimizer -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/nowait-reasoning-optimizer, .gemini/skills/nowait-reasoning-optimizer, .github/skills/nowait-reasoning-optimizer and .opencode/skills/nowait-reasoning-optimizer in your project.

What does Nowait Reasoning Optimizer need to run?

Going by SKILL.md and its folder, Nowait Reasoning Optimizer needs Python for the scripts in its folder. Our summary lists: Python 3.

Does Nowait Reasoning Optimizer access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Nowait Reasoning Optimizer safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Nowait Reasoning Optimizer use?

Nowait Reasoning Optimizer is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Nowait Reasoning Optimizer use?

About 1.2k tokens (SKILL.md is roughly 4.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Nowait Reasoning Optimizer?

Skills that share tags, products or a category with Nowait Reasoning Optimizer: LLM Council on Fireworks AI (dair-ai/dair-academy-plugins, 614 stars), Claude Maintain Models (Kiln-AI/Kiln, 5.2k stars), LLM Pipeline Profiler Analysis (BBuf/AI-Infra-Auto-Driven-SKILLS, 925 stars) and Update Ollama Cloud Models (heypinchy/pinchy, 182 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Nowait Reasoning Optimizer?

davila7 (a GitHub user) maintains it in davila7/claude-code-templates, which has 32,483 GitHub stars. The repository holds 478 skills in this directory. The repository was last updated on October 9, 2026.

Source: davila7/claude-code-templates on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.