How do I install the Tw Legal RAG skill?

Run npx skillmds@latest add aa0101181514/tw-legal-rag in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Tw Legal RAG safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: executes scripts. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Tw Legal RAG?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Tw Legal RAG free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Tw Legal RAG?

aa0101181514 (@aa0101181514) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Tw Legal RAG by aa0101181514 — Agent Skill

Name: Tw Legal RAG
Rating: 4.5 (4 reviews)
Author: aa0101181514

Tw Legal RAG

Retrieve real Taiwan court judgments with verifiable citations before answering any question about Taiwan law or case law. Use when the user asks about 台灣/臺灣 judgments, 判決, 法院見解, legal precedents, or wants case support for a legal argument. Wraps the twlegalrag CLI (pack subcommand); answers must cite only citation_ids present in the returned bundle.

aa0101181514 Updated Sep 22, 2026

File contents

TW Legal RAG — Taiwan judgment retrieval for Claude Code

When to use

Any time the question involves Taiwan law and the answer would benefit from actual court judgments: legal questions in plain language, case-law research, checking whether a cited 字號 exists, or building an argument that needs precedent support. Retrieve FIRST, answer second.

How to retrieve

Run the bundled script with the user's question:

python3 scripts/search_judgments.py "雇主未足額提撥勞工退休準備金的法律責任" -n 5

It prints a bundle JSON containing judgments (court, case number, date, reasoning excerpts), stable citation ids (J1, J2, ...), an allowed_citations whitelist and verification instructions.

The script locates the twlegalrag executable automatically and falls back to python -m twlegalrag when the console script is not on PATH (common on Windows user-site installs). If both fail, install the CLI first: pip install twlegalrag.

Citation rules (non-negotiable)

Cite ONLY ids listed in the bundle's allowed_citations. Judgments in unread_candidates were not read in full — never present them as authority.
Quote only text that appears verbatim in a judgment's fulltext_excerpt. Do not reconstruct holdings from memory.
If the bundle is empty or retrieval fails, say so explicitly. Never substitute invented case numbers.
A case number appearing in the bundle proves the judgment exists — it does not prove it supports the user's position. Read the excerpt before characterising a holding.
Follow the bundle's own verification_instructions field; it is part of the data contract.

Notes

Retrieval hits the public TLR endpoint (tlr.dr-legal.com.tw); no API key is required and no LLM is called server-side.
Windows: if output shows garbled Chinese, the terminal is not in UTF-8; the CLI reconfigures its own streams, but pipe output to a file when in doubt.

---name: tw-legal-ragdescription: Retrieve real Taiwan court judgments with verifiable citations before answering any question about Taiwan law or case law. Use when the user asks about 台灣/臺灣 judgments, 判決, 法院見解, legal precedents, or wants case support for a legal argument. Wraps the twlegalrag CLI (pack subcommand); answers must cite only citation_ids present in the returned bundle.---# TW Legal RAG — Taiwan judgment retrieval for Claude Code## When to useAny time the question involves Taiwan law and the answer would benefit fromactual court judgments: legal questions in plain language, case-law research,checking whether a cited 字號 exists, or building an argument that needsprecedent support. Retrieve FIRST, answer second.## How to retrieveRun the bundled script with the user's question:```bashpython3 scripts/search_judgments.py "雇主未足額提撥勞工退休準備金的法律責任" -n 5```It prints a bundle JSON containing judgments (court, case number, date,reasoning excerpts), stable citation ids (J1, J2, ...), an`allowed_citations` whitelist and verification instructions.The script locates the `twlegalrag` executable automatically and falls backto `python -m twlegalrag` when the console script is not on PATH (common onWindows user-site installs). If both fail, install the CLI first:`pip install twlegalrag`.## Citation rules (non-negotiable)1. Cite ONLY ids listed in the bundle's `allowed_citations`. Judgments in `unread_candidates` were not read in full — never present them as authority.2. Quote only text that appears verbatim in a judgment's `fulltext_excerpt`. Do not reconstruct holdings from memory.3. If the bundle is empty or retrieval fails, say so explicitly. Never substitute invented case numbers.4. A case number appearing in the bundle proves the judgment exists — it does not prove it supports the user's position. Read the excerpt before characterising a holding.5. Follow the bundle's own `verification_instructions` field; it is part of the data contract.## Notes- Retrieval hits the public TLR endpoint (`tlr.dr-legal.com.tw`); no API key is required and no LLM is called server-side.- Windows: if output shows garbled Chinese, the terminal is not in UTF-8; the CLI reconfigures its own streams, but pipe output to a file when in doubt.

Frequently asked questions