Reverse-engineer job descriptions: translate jargon ("5 years" → 3), infer company culture (red flags, vibes), match to your resume, detect negotiation signals. Outputs Markdown + shareable HTML report. Works with any OpenAI-compatible LLM (Ollama, DeepSeek, OpenAI, etc.). Supports text paste (primary), URL scraping (BOSS/拉勾/LinkedIn), and file input (.txt/.md/.docx/.pdf).

Install

openclaw skills install @ucsdzehualiu/jd-truth-detector

jd-truth-detector

Reverse-engineer job descriptions across 4 dimensions: jargon translation, company culture inference, resume match analysis, negotiation signals.

Quick Start

bash
pip install -r requirements.txt

# Analyze from text
python scripts/jd_analyze.py --jd-text "<paste JD here>"

# From URL (experimental)
python scripts/jd_analyze.py --jd-url "https://www.zhipin.com/job_detail/..."

# With resume
python scripts/jd_analyze.py --jd-file jd.txt --resume-file resume.md

Configuration

bash
export JTD_BASE_URL=http://localhost:11434/v1
export JTD_API_KEY=ollama
export JTD_MODEL=qwen2.5:72b

Or ~/.config/jd-truth-detector/config.yaml:

yaml
base_url: https://api.deepseek.com/v1
api_key: sk-...
model: deepseek-chat

Four Dimensions

DimensionWhat it does
黑话翻译"5年经验" → 实际3年即可
公司体感红旗词频率、技术成熟度、业务清晰度
简历匹配(可选) 对比 JD 和你的简历
议价信号薪资透明度、职级明确性、急招信号

Privacy

JD + resume sent only to your configured LLM endpoint. No telemetry.