os-control

os-control is a skill for Claude Code, Codex from Willian-2-0-0-1/os-control-mcp. It costs 29 tokens per session (675 once invoked), scanned A, original, MIT.

A ferramenta para manter sessões persistentes de Bash, CMD, PowerShell e WSL, incluindo trabalhos autónomos acompanhados até terminarem.

In plain words
What is it for?
Serve para executar comandos em diferentes shells, continuar trabalhos numa sessão existente e verificar se os critérios definidos foram cumpridos.
Why use it?
Evita perder o estado entre comandos e permite acompanhar um trabalho longo até ser concluído, bloqueado, falhar ou ser cancelado.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Serve para executar comandos em diferentes shells, continuar trabalhos numa sessão existente e verificar se os critérios definidos foram cumpridos.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/willian-2-0-0-1/os-control-mcp/os-control
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.

Any agent
npx skills add Willian-2-0-0-1/os-control-mcp --skill os-control
Clone the repo
git clone --depth 1 https://github.com/Willian-2-0-0-1/os-control-mcp

Made for: Claude Code, Codex.

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 os-control

README.md
[![agentmods](https://agentmods.dev/badge/skills/willian-2-0-0-1/os-control-mcp/os-control/github.svg)](https://agentmods.dev/skills/willian-2-0-0-1/os-control-mcp/os-control)
Your own site
<a href="https://agentmods.dev/skills/willian-2-0-0-1/os-control-mcp/os-control"><img src="https://agentmods.dev/badge/skills/willian-2-0-0-1/os-control-mcp/os-control/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.

agentmods 80×15 button for os-control

Your own site · 80×15
<a href="https://agentmods.dev/skills/willian-2-0-0-1/os-control-mcp/os-control"><img src="https://agentmods.dev/badge/skills/willian-2-0-0-1/os-control-mcp/os-control.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 29 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 675 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00029 $0.00675
Opus 5 $0.00015 $0.00338
Sonnet 5 $0.00006 $0.00135
Haiku 4.5 $0.00003 $0.00068

Measured 10d ago against content hash 846adc37c881, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

os-control 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 10d 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.

skills/os-control/SKILL.md · 61 lines

How it starts

The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.

OS Control MCP

Modo autônomo

Quando o usuário pedir para trabalhar até concluir, crie um objetivo com os_goal_start, usando autonomy: "autonomous", uma lista objetiva de acceptance_criteria, max_steps e timeout_seconds.

Depois, mantenha o ciclo abaixo até o objetivo ficar completed, blocked, failed ou cancelled:

  1. escolha uma ação pequena e in-scope;
  2. execute pela sessão vinculada usando goal_id;
  3. leia a saída nova com os_read_output;
  4. valide o resultado e registre evidências com os_goal_update;
  5. continue para next_action sem encerrar a conversa prematuramente.

Não marque completed apenas porque um comando retornou; comprove todos os critérios. Se não houver progresso após tentativas razoáveis, marque blocked com o motivo. Os limites do objetivo evitam loops infinitos, mas não são uma aprovação manual.

Sessões

Use os_list_sessions antes de operar uma sessão existente. Confirme o session_id, o shell e o diretório antes de enviar entrada.

O MCP suporta vários chats simultâneos. Quando o cliente fornecer um identificador estável de conversa, preserve-o como client_id em todas as chamadas para manter sessões e objetivos separados. Sem client_id, trate a lista como estado compartilhado do servidor e nunca assuma que uma sessão criada nesta conversa é a única sessão ativa.

Use os_send_input para digitar no terminal e os_read_output com o último next_seq e next_offset para buscar apenas a saída nova. Pode usar max_bytes grande quando a tarefa exigir saída extensa. Se a resposta vier com has_more: true ou truncated: true, continue imediatamente usando os cursores retornados; não descarte o restante nem marque a etapa como concluída. Não trate o retorno de os_send_input como prova de que o comando terminou; avalie a saída com os_read_output e o estado com os_session_status.

Em um objetivo autonomous, não peça confirmação a cada comando: o usuário autorizou a execução contínua dentro dos limites do objetivo. Fora de um objetivo autônomo, peça confirmação explícita antes de comandos destrutivos, alteração de rede, instalação de software, encerramento de processos ou encerramento de sessões. Nunca envie chaves, senhas ou tokens para o terminal.

Read the full file on GitHub · 61 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. 10d ago First seen · 61 lines · 29 tokens per session scan A 846adc37c881

Subscribe to this mod's changes

os-control is a skill published in the GitHub repository Willian-2-0-0-1/os-control-mcp (0 stars, last pushed 12d ago), licensed MIT. It adds 29 tokens to every session and 675 once invoked, about $0.0001 per session on Opus 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.