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.
npx skills add calneymgp/solodev --skill dev-plangit clone --depth 1 https://github.com/calneymgp/solodevWrote 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.
[](https://agentmods.dev/skills/calneymgp/solodev/dev-plan)<a href="https://agentmods.dev/skills/calneymgp/solodev/dev-plan"><img src="https://agentmods.dev/badge/skills/calneymgp/solodev/dev-plan/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/calneymgp/solodev/dev-plan"><img src="https://agentmods.dev/badge/skills/calneymgp/solodev/dev-plan.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00129 | $0.03076 |
| Opus 5.5 | $0.00052 | $0.01230 |
| Sonnet 5.5 | $0.00026 | $0.00615 |
| Haiku 4.5 | $0.00013 | $0.00308 |
Grade A, and why
dev-plan 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 1mo ago.
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.
How it starts
The opening of the file, as written. The whole thing — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/dev-plan — Plano atômico, reset-friendly, sem código
Esta skill produz .plans/<feature>/PLAN.md — um documento auto-suficiente. Uma sessão nova de Claude Code, lendo só esse arquivo + o CLAUDE.md do projeto, deve conseguir executar a feature inteira via /dev-coding.
Princípios não-negociáveis
-
Sem código no plano. Sem snippets de implementação, sem TypeScript, sem SQL. Apenas: decisões, áreas afetadas, contratos de interface (prose), critérios de aceite verificáveis. Exceção rara: snippet de tipo/schema/state-machine quando isso encoda a decisão de forma mais precisa que prosa.
-
Tasks atômicas com critério verificável. Cada task tem
acceptanceque pode ser checada via grep / build / test, não "feito quando funcionar". -
Vertical slices preferidos. Cada task corta TODAS as camadas (schema → API → UI → test) em uma fatia fina, NÃO uma camada inteira por vez. Antipadrão: task-01 = todos os models, task-02 = todas as APIs, task-03 = toda UI.
-
Reset protocol embutido. O plano contém a seção
## Reset Protocolno fim — instruções para sessão nova retomar do zero. -
Karpathy mínimo necessário. Não invente fases, milestones, OKRs, riscos especulativos. Se a feature cabe em 5 tasks, são 5 tasks — não invente sub-sprints.
-
Decisão sem evidência é chute. Toda decisão material registra escolha + alternativa descartada + motivo em 1 linha. Por quê: o executor (você pós-
/clear, ou outra sessão) não deve reinventar tradeoffs já feitos — nem descobrir tarde que havia uma opção melhor. -
Nada executa sem Approve. O plano é apresentado uma vez, o usuário responde
Approve/Changes/Cancel, e só então alguém executa. Por quê: corrigir granularidade no plano custa 1 turn; corrigir no meio do/dev-codingcusta a sessão.
Processo
1. Carregue contexto
Em ordem:
- Read
.plans/<feature>/BRIEF.mdse existir → fonte primária - Read
CLAUDE.mddo projeto + sub-CLAUDEs relevantes (ex.:src/app/CLAUDE.md) - Read glossário/schema/CONTEXT do projeto se aplicável
- Se não houver BRIEF, faça grilling rápido (3-5 perguntas críticas) — modo
/dev-brainstormcondensado
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 1mo ago Changed · +36 lines f8ecea17e5cc
- 1mo ago First seen · 149 lines · 129 tokens per session scan A 7d982880ff91
dev-plan is a skill published in the GitHub repository calneymgp/solodev (2 stars, last pushed 1mo ago), licensed MIT. It adds 129 tokens to every session and 3,076 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
ai-engineering-toolkit
6 production-ready AI engineering workflows: prompt evaluation (8-dimension scoring), context budget planning, RAG pipeline design, agent security audit (65-point checklist), eval harness building, and product sense coaching.
ai-prompt-regression-testing
Prompt engineering regression test matrix register: baseline outputs, semantic drift thresholds, judge evaluations, and golden dataset.
q-thinking-chain
Use this skill when the user needs to think deeply, stress test an idea, make an important decision, challenge their own thinking, or wants to go beyond a surface level answer. Activate for analysis, critique, decision support, or deep reasoning tasks.
q-thinking
Use this skill when the user needs to understand a problem before solving it. Activate for complex situations where the answer is not clear, where assumptions need surfacing, or where the user feels stuck or overwhelmed. This skill helps map what is actually happening before any action is taken.
q-building
Use this skill when the user is setting up a workflow, an agent, a research process, or a plan they want to actually run. Activate when the user wants structure around how work gets done rather than help with a single one-off task, or when a repeating process keeps producing inconsistent results.
q-claude-setup
Use this skill when the user wants to configure Claude to know them deeply and improve itself over time. Activate when setting up projects, context files, or self-improving instructions, or when the user wants Claude to compound across sessions instead of starting from zero each time.