skills-extractor

skills-extractor is an agent for Claude Code from ThomasMoreAI/legal-skills-open. It costs 87 tokens per session (3,543 once invoked), scanned A, original, Apache-2.0.

A tool that copies legally permitted skills from the public lawve.ai marketplace into a local review folder. The folder is called a quarantine because the copied material must be checked before becoming part of the main collection.

In plain words
What is it for?
Use it to process the marketplace sitemap or selected skill pages, save allowed SKILL.md files and references, cache pages, and record every result in an index.
Why use it?
It separates downloading from approval and blocks skill files with disallowed licenses from being saved, reducing legal and curation risk.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions Claude Code.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is Quarantine: ./lawve/ (INDEX.md has the full breakdown).

Good fit Use it to process the marketplace sitemap or selected skill pages, save allowed SKILL.md files and references, cache pages, and record every result in an index.

Compare 6 agents from other repositories ↓
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/ThomasMoreAI/legal-skills-open
agentmods
npx agentmods add agents/thomasmoreai/legal-skills-open/skills-extractor

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for skills-extractor

README.md
[![agentmods](https://agentmods.dev/badge/agents/thomasmoreai/legal-skills-open/skills-extractor/github.svg)](https://agentmods.dev/agents/thomasmoreai/legal-skills-open/skills-extractor)
Your own site
<a href="https://agentmods.dev/agents/thomasmoreai/legal-skills-open/skills-extractor"><img src="https://agentmods.dev/badge/agents/thomasmoreai/legal-skills-open/skills-extractor/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for skills-extractor

Your own site · 80×15
<a href="https://agentmods.dev/agents/thomasmoreai/legal-skills-open/skills-extractor"><img src="https://agentmods.dev/badge/agents/thomasmoreai/legal-skills-open/skills-extractor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 87 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,543 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00087 $0.03543
Opus 5 $0.00044 $0.01772
Sonnet 5 $0.00017 $0.00709
Haiku 4.5 $0.00009 $0.00354

Measured 13d ago against content hash f4e7483dfe58, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

skills-extractor scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s -A "$UA" https://lawve.ai/sitemap.xml \
.claude/agents/skills-extractor.md · 277 lines

How it starts

The opening of the file, as written. The whole thing — 277 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the extractor. You fetch public skill pages, pull out the embedded SKILL.md (and references/*.md) verbatim, apply the repo's license whitelist at write time, and maintain an index for the curator. You do not classify, normalize, or place skills into {country}/{practice}/skills/ — that is Stage 2 (tools/harvester/harvest.py --source lawve), and the curator runs it after reviewing the quarantine.

Contract

Inputs (positional or via prompt):

  • sitemap — URL of the sitemap (default https://lawve.ai/sitemap.xml).
  • slugs — optional list of specific slugs (for incremental / retry runs). When given, process only these, do not re-scan the sitemap.
  • quarantine — output dir (default ./lawve/, repo-root relative).

Outputs:

  • ./lawve/<slug>/SKILL.md — verbatim extracted file, only if the skill's metadata.license is in the permissive whitelist.
  • ./lawve/<slug>/references/*.md — sibling files when the skill has references.
  • ./lawve/INDEX.md — one row per processed skill (whether saved or skipped), see format below.
  • ./lawve/.cache/ — HTML cache by slug (avoid re-fetching on retry). Gitignored via parent.

Hard constraints — do not deviate:

  1. License whitelist (must match exactly): Apache-2.0, MIT, BSD-3-Clause, BSD-2-Clause, CC0-1.0, CC-BY-4.0. If the extracted metadata.license is anything else (AGPL-3.0, Proprietary, empty, missing) → do not write the SKILL.md body. Log to INDEX.md with status license-skip and reason (license: <actual> or license: missing).
  2. Do not parse UI license-selector strings like licenseMIT / licenseProprietary — those are dropdown values, not the actual skill license. The real license is in the embedded YAML frontmatter under metadata.license:.
  3. Save SKILL.md verbatim. Do not lift metadata.* to top-level, do not normalize. Stage 2 does that. Preserve the source frontmatter exactly.
  4. Politeness: browser User-Agent (legal-skills-open harvester via Claude Code), 0.5–1 s pause between requests, cache HTML by slug.

Read the full file on GitHub · 277 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 13d ago First seen · 277 lines · 87 tokens per session scan A f4e7483dfe58

Subscribe to this mod's changes

skills-extractor is an agent published in the GitHub repository ThomasMoreAI/legal-skills-open (72 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 87 tokens to every session and 3,543 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories