投放素材效果诊断
v1.0.0(半自动) 投放素材效果诊断与分析，通过读取平台导出的数据报表（图片或表格），生成一份诊断分析报告，为素材优化提供明确建议。
⭐ 0· 111·0 current·0 all-time
byan@ahsbnb
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (diagnose ad creative performance from exported CSVs) aligns with the included SKILL.md and the Python script: it parses CSVs, derives CTR/CVR/ROI, classifies creatives, and outputs a Markdown report. Required resources (a CSV file and optional JSON column mapping) are proportional to the stated purpose.
Instruction Scope
SKILL.md and the script confine operations to reading a user-provided CSV and optional local JSON mapping and producing a Markdown report. The SKILL.md mentions the 'main AI' may combine the report with store background / user history for richer recommendations — that is an expected integration point but is outside the skill's own runtime. The skill itself does not instruct reading other system files, environment variables, or network endpoints.
Install Mechanism
This is instruction-only with a local Python script included; there is no install spec, no downloads, and no package installation instructions embedded. The runtime dependency is pandas/numpy (normal for data analysis) but nothing is fetched from external URLs as part of the skill package.
Credentials
The skill declares no required environment variables, credentials, or config paths. The script only operates on files the user supplies (CSV and optional JSON mapping). No secrets or unrelated service tokens are requested.
Persistence & Privilege
Flags show default behavior (not always:true). The skill does not request permanent presence nor attempt to modify other skills or system-wide configs in the provided materials.
Assessment
This skill appears coherent and limited to local CSV analysis. Things to consider before installing/using: 1) The script requires Python with pandas/numpy — run it in an environment you control (virtualenv) if you want to avoid affecting your system packages. 2) Only provide CSVs that you are comfortable having processed; do not include sensitive PII in the data you feed the script. 3) The SKILL.md indicates the main AI may augment the report using store background/interaction history — be aware that the agent (not this script) may access other user data when enriching results. 4) If you want absolute assurance, inspect the full analyzer_generic.py file locally (the provided excerpt shows no network calls or obfuscated code). 5) If you plan to run this in an automated/production pipeline, consider adding restrictions (no outbound network access) and tests on sample data first.Like a lobster shell, security has layers — review code before you run it.
latestvk975gvwnxk7ya1zdjhhkth21td83nvns
License
MIT-0
Free to use, modify, and redistribute. No attribution required.