by 安全扫描
OpenClaw
可疑
medium confidenceThe skill is an instruction-only, migrated ad-creative example that is internally plausible but provides no actionable implementation and references absolute local skill paths (~/.opencode/skills), which is inconsistent and could lead to unexpected local filesystem access if executed.
评估建议
This SKILL.md appears to be a migrated example, not a complete, self-contained implementation. Before using or installing: (1) inspect any actual code that will be executed — the SKILL.md references modules under ~/.opencode/skills which would access your home directory; (2) ensure the referenced modules exist and review them for unexpected behavior; (3) do not run unknown code in a production environment — test in a sandbox; (4) request or verify a concrete implementation or an install spec if ...详细分析 ▾
ℹ 用途与能力
The name/description claim a full-featured ad-creative skill, but the SKILL.md contains only metadata and example snippets rather than concrete runtime instructions. The examples reference other local skill modules (e.g., ~/.opencode/skills/subagente-optimizado, buscador_de_skills, exa-search), which is plausible for integration but not justified or explained in the document — this makes the claimed capabilities unclear and dependent on external local components that are not provided.
⚠ 指令范围
SKILL.md is effectively a placeholder with usage examples that require loading modules from the user's home directory (~/.opencode/skills). While the file itself doesn't instruct the agent to read arbitrary files, the example code implies the skill (or consumers of it) will require local filesystem modules. That introduces the possibility of unexpected local file access or dependency on other local skills/subagents not present here.
✓ 安装机制
There is no install spec and no code files beyond SKILL.md, so nothing is written to disk by an installer. This is low-risk from an installation perspective.
✓ 凭证需求
The skill declares no required environment variables, credentials, or config paths. There are no direct requests for sensitive tokens or unrelated credentials.
✓ 持久化与权限
always is false and the skill does not request persistent presence or system-wide configuration changes. It does not declare autonomous privileges beyond the platform defaults.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv1.0.02026/3/30
ad-creative-migrated 1.0.0 - Primera versión migrada desde ./examples con capacidades completas. - Documentación actualizada y optimizada para Qwen Code. - Añadidas directrices de integración y ejemplos de uso básico/avanzado. - Implementadas optimizaciones: compresión de contexto, restricción de tools, límites de iteración y caché de resultados. - Etiquetas actualizadas para reflejar migración y funcionalidad.
● 无害
安装命令
点击复制官方npx clawhub@latest install ad-creative-migrated
镜像加速npx clawhub@latest install ad-creative-migrated --registry https://cn.longxiaskill.com
技能文档
Overview
Esta skill ha sido migrada desde ./examples y mejorada para Qwen Code.
Usage Examples
Example 1: Basic Usage
const skill = require('~/.opencode/skills/ad-creative');
const result = await skill.execute();
Example 2: Advanced Usage
const { OptimizedSubagent } = require('~/.opencode/skills/subagente-optimizado');const agent = new OptimizedSubagent({ taskName: 'ad-creative task', taskType: 'ad', context: { userGoal: 'Objective' } });
const result = await agent.execute(async ({ context, tools }) => { // Implementation });
Integration
Buscador de Skills
const bs = require('~/.opencode/skills/buscador_de_skills');
const skill = bs.searchByName('ad-creative');
Exa Search
const search = require('~/.opencode/skills/exa-search');
const results = await search.searchWeb('ad-creative best practices 2026');
Optimization
Esta skill usa las optimizaciones de quota:
- Context compression (650 tokens vs 10K)
- Tool restriction (3-4 tools max)
- Iteration limits (3 max)
- Result caching (TTL: 1h)
References
- Source:
./examples/claude-skills-main/.gemini/skills/ad-creative - Migrated: 2026-03-29