dev-fix

A structured debugging guide for fixing software bugs through a reproducible feedback loop. It first classifies the bug, tests likely causes, adds a regression test, and removes temporary diagnostic work.

In plain words
What is it for?
Use it to investigate unclear or intermittent bugs, apply a minimal fix, verify the failing command, run relevant tests, and record the cause.
Why use it?
It reduces guesswork by requiring evidence before changing code. It also separates simple bugs from problems that need broader architectural decisions.

Skill for Claude CodeCodex

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 skills/calneymgp/solodev/dev-fix
Any agent
npx skills add calneymgp/solodev --skill dev-fix
Clone the repo
git clone --depth 1 https://github.com/calneymgp/solodev

Made for: Claude Code, Codex.

Per session 124 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,173 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. 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 $0.00124 $0.01173
Opus 5 $0.00062 $0.00587
Sonnet 5 $0.00025 $0.00235
Haiku 4.5 $0.00012 $0.00117

Measured yesterday against content hash 345309f1f390, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dev-fix scanned grade A with 1 finding 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

O teste/script que reproduz o bug em **<10s** com pass/fail determinístico. **Isso é 90% do trabalho.** Se o seu único loop é "rodar o app e clicar", invista em melhorá-lo antes de qualquer hipótese: um teste, um curl, u
skills/dev-fix/SKILL.md · 80 lines

How it starts

The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/dev-fix — Bug morre com método, não com tentativa

Bugfix não precisa de BRIEF nem PLAN. Precisa de disciplina de diagnóstico — a diferença entre 10 minutos e 3 horas é quase sempre a qualidade do feedback loop, não a dificuldade do bug.

Triagem (primeiro turn)

Classifique em 1 linha e siga o caminho:

Classe Heurística Caminho
Trivial Causa evidente no erro (typo, import, null óbvio), fix de 1-5 linhas Modo rápido
Real Causa não evidente, comportamento intermitente, "funcionava ontem" Loop completo
Arquitetural O fix exigiria redesenho, toca contrato público ou 3+ módulos Pare. Apresente diagnóstico + opções. Sugira /dev-brainstorm ou /dev-plan.

Modo rápido (bug trivial)

  1. Leia o arquivo inteiro em volta do erro (não só a linha)
  2. Aplique o fix mínimo
  3. Rode o comando que falhava → veja passar
  4. Se existir suite, rode os testes do módulo tocado
  5. Reporte em 2 linhas: causa → fix

Se o "trivial" não morrer no primeiro fix, ele não era trivial. Promova para o loop completo — não tente um segundo palpite.

Loop completo (6 fases)

Phase 1 — Construa o feedback loop

O teste/script que reproduz o bug em <10s com pass/fail determinístico. Isso é 90% do trabalho. Se o seu único loop é "rodar o app e clicar", invista em melhorá-lo antes de qualquer hipótese: um teste, um curl, um script de 5 linhas.

Phase 2 — Reproduza

Rode o loop, observe a falha, confirme que é a mesma que o usuário descreveu (não uma vizinha). Se não reproduz: colete mais contexto (env, dados, versão) antes de teorizar.

Phase 3 — Hipóteses ranqueadas

3-5 hipóteses, cada uma com predição falsificável ("se X é a causa, mexer em Y faz desaparecer"). Mostre a lista ao usuário antes de testar — ele ranqueia mais rápido com domain knowledge. Comece pela mais provável OU pela mais barata de descartar.

Phase 4 — Instrumente

Um probe por hipótese. Prefira debugger > log direcionado > log everything. Logs de debug com prefixo único ([DEBUG-a4f2]) para cleanup via grep no fim. Cada probe responde sim/não para UMA hipótese — probe que "olha geral" é ruído.

Read the full file on GitHub · 80 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 · 80 lines · 124 tokens per session scan A 345309f1f390

Subscribe to this mod's changes

dev-fix is a skill published in the GitHub repository calneymgp/solodev (2 stars, last pushed 2mo ago), licensed MIT. It adds 124 tokens to every session and 1,173 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

karpathy-guidelines

Tool-agnostic behavioral guidelines for AI coding assistants. Use when writing, reviewing, debugging, or refactoring code to reduce overengineering, surface ambiguity, make surgical changes, and define verifiable success criteria.

Vincent-A-Yang/karpathy-skills-anycoding · 47 tokens

Evolve-Skill

核心进化技能。在开发结束时调用,自动分析对话历史,将“项目进化资产”沉淀到 EVOLVE.md,并将平台特有的 AI 行为教训写入 CLAUDE.md / GEMINI.md / AGENTS.md / CURSOR.md,实现可审计、可复用的持续进化。触发词:"总结经验"、"进化"、"evolve"、"复盘"、"summarize lessons"、"retrospective"、"postmortem"。不适用于仅完成简单查询、单文件小改动或未产生可沉淀资产的场景。.

Ecrypted-Data/Evolve-Skill · 143 tokens

metrics-instrumentation

Specification for instrumenting an opik-backend workflow with operational OpenTelemetry metrics — per-stage throughput/latency/error counters and native histograms, dimensioned per-customer (workspace). Use when a pipeline (scoring, ingestion, experiments, jobs) needs per-stage visibility. Covers metric emission only…

comet-ml/opik · 93 tokens

happiness-skill

当用户问「怎么才能更幸福/为什么得到了还不满足/怎么减少焦虑」时调用。 核心理念: 幸福是缺憾感清空的默认状态, 是可训练的技能; 欲望是与自己的契约(得到前不快乐), 同时只留一个重大欲望; 活在当下。 不适用于: 临床抑郁等需要专业治疗的场景(本书方法不能替代医疗)。 Triggers: 幸福/不快乐/欲望/焦虑/知足/活在当下/happiness/desire/anxiety.

kangarooking/cangjie-skill · 136 tokens

api-design

REST/GraphQL/gRPC API design best practices. Use when designing APIs, defining contracts, handling versioning. Covers OpenAPI 3.2, GraphQL Federation, gRPC streaming.

majiayu000/spellbook · 42 tokens

setup-matt-pocock-skills

为本仓库配置工程技能——设置其 issue tracker、分诊标签词汇表和领域文档布局。首次使用其他工程技能前运行一次。.

devcxl/mattpocock-skills-zh · 43 tokens