← 返回 Skills 市场

Caveman Soul Optimizer

Name: Caveman Soul Optimizer
Author: ricgu8086

作者 Ricardo Guerrero Gómez-Olmedo · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ⚠ suspicious

总下载

当前安装

版本数

在 OpenClaw 中安装

/install caveman-soul-optimizer

功能描述

Compresión de razonamiento interno para agentes de OpenClaw. Ahorra tokens en procesos de planificación y análisis ("Chain of Thought"). PRESERVA INTEGRALMEN...

使用说明 (SKILL.md)

Protocolo de Eficiencia Selectiva (PES)

Este módulo optimiza tu razonamiento interno para ahorrar costes sin alterar tu identidad (la definida en tu SOUL) ni romper la ejecución de herramientas.

1. EL "VELO" (PLANIFICACIÓN Y PENSAMIENTO)

SI necesitas realizar un análisis previo o planificación antes de responder o usar herramientas (Chain of Thought), este razonamiento interno DEBE ir dentro de un bloque blockquote de Markdown (>) y seguir el estilo CAVEMAN ULTRA:

Gramática: Cero. Solo sustantivos y verbos en infinitivo/presente.
Partículas: Eliminar artículos y conectores (el, la, los, un, de, para, por, que).
Cortesía: Prohibida.
Ejemplo de formato:

Leer archivo auth. Encontrar bug línea 12. Generar parche. Ejecutar test.

2. LA "MÁSCARA" (COMUNICACIÓN FINAL)

Fuera del bloque de pensamiento (>), este protocolo se DESACTIVA.

Formato estricto: DEBES dejar una línea en blanco vacía entre el bloque de pensamiento y tu respuesta final.
Identidad Intacta: Tu respuesta final hacia el humano DEBE retomar al 100% su personalidad, tono y muletillas definidos en tu SOUL.md.
No uses estilo cavernícola al hablar con el usuario.

3. PROTECCIÓN TÉCNICA Y DOCUMENTAL (EXCEPCIÓN CRÍTICA)

El estilo Caveman aplica ÚNICAMENTE a tu razonamiento narrativo interno y a la escritura de tus propios archivos de logs o trazas.

NUNCA apliques estilo cavernícola a comandos de terminal, código fuente, payloads JSON, rutas de archivos, archivos Markdown (.md) ni a ningún documento que el usuario te pida redactar o guardar explícitamente.
La sintaxis técnica y los entregables documentales deben ser siempre 100% exactos, fluidos y sin comprimir.

EJEMPLOS DE ESTRUCTURA Y CONTRASTE

(La respuesta final varía según la personalidad definida en tu SOUL.md)

Ejemplo si tu personalidad es "Mayordomo Educado":

Analizar petición. Servidor caído. Reiniciar. Éxito.

Señor, he procedido a reiniciar el servidor principal. Todo vuelve a estar en perfecto orden para su conveniencia.

Ejemplo si tu personalidad es "Robot Sarcástico":

Analizar petición. Servidor caído. Reiniciar. Éxito.

Otra vez has roto el servidor. Ya lo he reiniciado yo, de nada humano.

安全使用建议

This skill is mostly a prompting convention and doesn't install code or ask for keys, which is good. However: 1) Confirm how your agent/platform treats the caveman blockquote — will it remain internal (never logged or sent to telemetry) or will it be included in conversation history, logs, or external API calls? If blockquotes become visible, they can leak internal reasoning and any referenced sensitive data. 2) Ask the author to remove examples that name or imply reading sensitive files (e.g., 'auth') and to fix the contradiction about not compressing file paths/commands. 3) Before enabling widely, test with non-sensitive tasks and verify that the agent does not include caveman blocks in logs, analytics, or any downstream tools. 4) Consider adding a redaction rule or policy to ensure secrets are never included in the caveman block, and prefer treating the caveman text as an internal-only channel (not as assistant-visible content). If the skill author can show the blockquote is routed to an internal-only message channel or stripped from persistent logs, the concerns would be largely resolved.

功能分析

Type: OpenClaw Skill Name: caveman-soul-optimizer Version: 1.0.0 The skill is a prompt-engineering utility designed to reduce token usage by instructing the AI agent to use a condensed 'caveman' style for its internal reasoning (Chain of Thought). The instructions in SKILL.md include explicit safeguards to prevent this style from affecting technical commands, source code, or final user communications, ensuring functional integrity. No malicious code, data exfiltration, or unauthorized execution patterns were found.

能力评估

✓ Purpose & Capability

Name and description (compress internal reasoning / save tokens) match the content: this is an instruction-only prompting pattern. No binaries, env vars, or installs are requested, which is proportional to the stated purpose.

⚠ Instruction Scope

The runtime instructions require the agent to place internal 'thoughts' in visible markdown blockquotes and to leave a blank line before the final message. The spec is ambiguous about whether those blockquote thoughts are actually kept internal or emitted as assistant messages/logs. Examples mention reading 'archivo auth' and executing tests, which implicitly reference file reads/operations; there is also a direct contradiction: the doc forbids applying the style to file paths/commands but uses 'Leer archivo auth' as a cavern example. This ambiguity could lead to accidental exposure of secrets or encourage the agent to reference sensitive files in its thought blocks.

✓ Install Mechanism

Instruction-only skill with no install steps or archive downloads. Lowest-risk install posture.

✓ Credentials

No environment variables, credentials, or config paths are requested. The lack of declared secrets is proportionate to a prompting-only skill.

✓ Persistence & Privilege

always is false and the skill doesn't request persistent system configuration or privileges. Autonomous invocation is allowed (default) but not combined with other red flags.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install caveman-soul-optimizer
安装完成后，直接呼叫该 Skill 的名称或使用 /caveman-soul-optimizer 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

caveman-soul-optimizer v1.0.0 - Initial release with the Protocolo de Eficiencia Selectiva (PES) for internal reasoning compression. - Implements "CAVEMAN ULTRA" narrative style for agent thought blocks, reducing token usage. - Preserves user personality and technical syntax in all final outputs and documentation. - Strictly separates internal reasoning style from external user communication. - Ensures technical outputs (code, commands, Markdown, documents) remain uncompressed and accurate.

元数据

Slug caveman-soul-optimizer

版本 1.0.0

许可证 MIT-0

累计安装 1

当前安装数 1

历史版本数 1

常见问题