技能详情（站内镜像，无评论）
作者：Adarsh More @AdarshVMore
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 0 · 167 · 1 current installs · 1 all-time installs
⭐ 0
安装量（当前） 1
🛡 VirusTotal ：良性 · OpenClaw ：可疑
Package：adarshvmore/report-generator-adarsh
安全扫描（ClawHub）
- VirusTotal ：良性
- OpenClaw ：可疑
OpenClaw 评估
The skill's instructions clearly require an OpenAI API key and reference other modules and logging, but the registry metadata claims no required credentials or install steps — this mismatch and the vague logging/fallback behavior warrant caution before installing.
目的
The declared purpose (single GPT-4.1-mini call to produce a marketing audit) matches the SKILL.md implementation pattern, but the registry metadata lists no required environment variables or primary credential while SKILL.md explicitly lists OPENAI_API_KEY as the auth dependency — that's an incoherence in declared requirements.
说明范围
SKILL.md is instruction-only and limited to: one system prompt, one user message with AuditData JSON, a single OpenAI call, parsing markdown, and returning a fallback if the API fails. It does not instruct reading system files or paths. However, it calls for token-usage and error logging with unspecified storage/targets and mentions a 'competitor collector fallback' (other modules) which are not included — these areas are underspecified and co…
安装机制
There is no install spec and no code files; being instruction-only is low-risk from an installation perspective.
证书
The skill requires an OpenAI API key (OPENAI_API_KEY) per SKILL.md but the registry lists no required env vars or primary credential. Requiring a secret API key is expected for this functionality, but it must be declared explicitly in metadata. Also, the SKILL.md states raw AuditData is sent to OpenAI; if that data contains PII or sensitive info, the credential and data exfiltration risk should be clearly documented.
持久
always:false and no install actions mean the skill does not request persistent system privileges. Autonomous invocation is allowed (default) but is not combined with other high-risk indicators here.
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「Report Generator Adarsh」。简介：Generates a structured marketing audit report from aggregated data using a sing…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/adarshvmore/report-generator-adarsh/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
# Report Generator Skill
## Purpose
Single GPT-4.1-mini call that transforms aggregated marketing data into a structured, professional audit report. This is the ONLY AI call in the entire audit pipeline.
## Input Schema
```typescript
interface AuditData {
input: AuditInput;
instagram: InstagramData;
metaAds: MetaAdsData;
keywords: KeywordData;
competitors: CompetitorData;
websiteAudit: WebsiteAuditData;
collectedAt: string;
}
```
## Output Schema
```typescript
interface MarketingReport {
brand: string;
generatedAt: string;
sections: ReportSections;
rawData: AuditData;
reportMarkdown: string;
}
```
## API Dependencies
- **API:** OpenAI
- **Model:** `gpt-4.1-mini`
- **Auth:** `OPENAI_API_KEY`
- **Cost:** ~$0.001-0.002 per call
## Implementation Pattern
- System prompt defines the analyst persona and exact 6-section format
- User message is the full AuditData JSON
- Single API call with `max_tokens: 1500`, `temperature: 0.4`
- Response markdown is parsed into individual sections via `## ` header splitting
- Token usage is logged for cost tracking
- Fallback report is generated if OpenAI call fails
## Token Budget
- Input: ~1,500-2,000 tokens (JSON data)
- Output: ~800-1,200 tokens (report)
## Error Handling
- Missing API key: returns fallback report with error message
- API failure: returns fallback report with raw data preserved
- All errors logged with context
## Example Usage
```typescript
const report = await generateMarketingReport(auditData);
console.log(report.reportMarkdown);
```
## Notes
- This is the ONLY file that should make OpenAI calls (except competitor collector fallback)
- Never add additional GPT calls to other modules