klotzkette/arbeitszeugnispruefer-skill — explained in plain English
Analysis updated 2026-05-18
Paste the SKILL.md instructions and an employment reference into an AI chat to get a graded assessment.
Identify sentences in a German employment reference that carry hidden negative meaning.
Draft a demand letter or litigation outline for challenging an unfavorable reference.
Test the skill against ten sample reference letters before using it on real documents.
| klotzkette/arbeitszeugnispruefer-skill | 0petru/sentimo | 0xblackash/cve-2026-46333 | |
|---|---|---|---|
| Stars | 17 | 17 | 17 |
| Language | — | Python | C |
| Setup difficulty | easy | moderate | moderate |
| Complexity | 1/5 | 3/5 | 4/5 |
| Audience | general | developer | researcher |
Figures from each repo's GitHub metadata at analysis time.
Not legal advice, the README raises unresolved questions about confidentiality and data protection law.
This repository contains a single structured text file designed to turn any capable AI chatbot into a tool for reviewing German employment references, known as Arbeitszeugnisse. Employment references in Germany follow strict conventions where positive-sounding phrases can carry hidden negative signals, and lawyers often need to decode these documents for clients who want to challenge a bad reference in court. The file bundles all the analysis instructions into one place so a law firm can use it without installing software or creating accounts. The core file is called SKILL.md. You paste its contents into a chat session with an AI model such as Claude, ChatGPT, or Gemini, then paste in the employment reference you want to review. The AI reads the instructions and produces a structured assessment using a traffic-light system, where each sentence gets marked red, orange, or green depending on how favorable or damaging its phrasing is under German legal standards. The output includes an overall grade estimate, a summary of findings, and suggestions for next steps. The eight-stage workflow runs from intake and identifying the type of reference, through sentence-by-sentence grading, detection of suspicious gaps or contradictions, up to drafting demand letters and outlining litigation strategy if the employee wants to sue for a corrected reference. The file references specific decisions from Germany's Federal Labor Court (BAG) to anchor the analysis in verified case law rather than guessing. The repository also includes ten sample reference letters across ten different industries, from pharmacy to logistics to nursing, which can be used to test or calibrate the skill before using it on real client documents. The project consolidates what was originally fifty separate plugin files into this single SKILL.md. The README includes a prominent disclaimer: the file is a technical experiment, not legal advice, and it makes no statement about whether using an AI service to process confidential client documents is lawful under German professional secrecy rules, data protection law, or the EU AI Act. Each user must resolve those questions independently before using it in practice.
A single instruction file that turns an AI chatbot into a tool for analyzing German employment references for hidden negative signals.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.