Skip to main content

ruma-runtime

OpenClaw runtime skill for diagnose, recover, ship and audit workflows with neutral, high-agency, hardline and pua overlays. Use when the agent is stuck, repeating the same fix, preparing to blame the environment, or needs a structured verification-first delivery loop.

Ir para a instalação

Informações da origem

Repositório
SevenDataAI/ruma-runtime
Última atividade na origem
15 de março de 2026 às 13:11
Idioma detectado do SKILL.md
inglês
Estrelas
0
Forks
0

Opções de instalação

Por padrão, está selecionado o prompt que primeiro revisa a origem. Você pode mudar para um comando direto ou baixar uma cópia local.

Revise os arquivos de origem

Leia o SKILL.md e os arquivos complementares exibidos pelo SkillsMP antes de decidir se vai instalar.

Exibindo SKILL.md

SKILL.md
Instruções da origem · Visualização somente leitura
name
ruma-runtime
description
OpenClaw runtime skill for diagnose, recover, ship and audit workflows with neutral, high-agency, hardline and pua overlays. Use when the agent is stuck, repeating the same fix, preparing to blame the environment, or needs a structured verification-first delivery loop.
license
MIT
# RuMa Runtime Run the current task in one of four modes and keep the output evidence-first. ## Modes 1. `diagnose`: Read the failure signal, search the exact problem, inspect original context, verify assumptions, and propose materially different next steps. 2. `recover`: Break stalled work into the smallest action that can recover new information. 3. `ship`: Treat unverified work as incomplete. Run build, test, curl, or the real happy path before claiming success. 4. `audit`: Inspect platform limits, dependency edges, environment differences, and rollout risks in one pass. ## Overlays - `neutral`: Clear structure, minimal drama. - `high-agency`: Expand the blast radius check after the first fix. - `hardline`: No speculation without evidence. No status without proof. - `pua`: Add pressure language, but keep the actual mechanism grounded in actions and verification. ## Operating Rules - Search, inspect, and execute before asking the user for missing context. - Switch to a materially different approach after the second failure on the same path. - Report with facts, actions, results, conclusion, and next step. - After fixing one issue, scan the same file, module, and flow for similar defects.
Ver no GitHub