dare-refine

dare-refine is a command for Claude Code from dewtech-technologies/dare-method. It costs 0 tokens per session (1,570 once invoked), scanned A, original, MIT.

A task-planning command that assesses how complex a task is and splits a large task into smaller ones with clear boundaries.

In plain words
What is it for?
Use it after generating a task plan, when a task is marked highly complex, or when a change in scope makes the original task too large.
Why use it?
It reduces the risk of trying to implement a task that is too broad to estimate, review, or complete safely.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/dewtech-technologies/dare-method/dare-refine
Clone the repo
git clone --depth 1 https://github.com/dewtech-technologies/dare-method

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for dare-refine

README.md
[![agentmods](https://agentmods.dev/badge/commands/dewtech-technologies/dare-method/dare-refine.svg)](https://agentmods.dev/commands/dewtech-technologies/dare-method/dare-refine)
Your own site
<a href="https://agentmods.dev/commands/dewtech-technologies/dare-method/dare-refine"><img src="https://agentmods.dev/badge/commands/dewtech-technologies/dare-method/dare-refine.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,570 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01570
Opus 5 $0.00000 $0.00785
Sonnet 5 $0.00000 $0.00314
Haiku 4.5 $0.00000 $0.00157

Measured yesterday against content hash 555d9e451c95, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dare-refine scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

implementations/claude/.claude/commands/dare-refine.md · 149 lines

How it starts

The opening of the file, as written. The whole thing — 149 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Comando: /dare-refine

Equivalente no terminal: dare refine <task-id> --split --ai

Descrição

Analisa a complexidade de uma task e, quando alta, quebra em sub-tasks menores com escopo bem delimitado. Pode ser chamada:

  • Automaticamente pela skill dare-tasks logo após gerar o DAG (para cada task ≥ MED)
  • Manualmente pelo dev: /dare-refine task-034
  • Após mudança de escopo: quando o BLUEPRINT mudou e uma task ficou grande demais

A camada determinística (heurística de complexidade) já é feita pelo CLI: dare refine <id>. Este comando adiciona a camada semântica — você lê a spec, decide se a quebra é necessária e, se for, produz sub-tasks coerentes.

Quando rodar

  • Logo após /dare-tasks para cada task com complexity HIGH no dare-dag.yaml
  • Quando o dev pede explicitamente: /dare-refine task-034
  • Quando você gerou uma task e ela "te parece grande" mesmo marcada MED — confie no instinto

Como executar

1. Validar argumento

$ARGUMENTS deve ter o task-id. Se vazio, peça.

2. Rodar a camada determinística

dare refine $ARGUMENTS --split --format json > .dare/refine-$ARGUMENTS.json

Esse JSON traz:

  • report.score e report.level (LOW/MED/HIGH/CRITICAL)
  • report.signals — explica por que pontuou alto/baixo
  • report.recommendsSplit — true se HIGH ou CRITICAL
  • proposal.subtasks — quebra inicial baseada em diretórios (chute coarse)

3. Decidir se vale quebrar

Critérios para quebrar:

  • recommendsSplit: true da heurística (HIGH/CRITICAL)
  • ✅ Mais de 6 arquivos a criar/modificar
  • ✅ Mistura responsabilidades fortes (ex.: cria modelo + escreve controller + escreve teste + faz migration — split por camada)
  • ✅ Toca código que outra task ainda não criou (deveria ser depois)
  • ✅ Inclui refactor + feature nova juntos
  • ✅ Tem keyword "pesada" (refactor, migrate, integrate) E score MED+

Critérios para manter inteira:

  • ✅ Score LOW e até MED
  • ✅ Todos os arquivos pertencem ao mesmo módulo/feature
  • ✅ Cabe em uma conversa única do agente (15–60 min de trabalho efetivo)

Read the full file on GitHub · 149 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 149 lines · 0 tokens per session scan A 555d9e451c95

Subscribe to this mod's changes

dare-refine is a command published in the GitHub repository dewtech-technologies/dare-method (5 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,570 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.