tax-doc-classifier — security grade SAFE, quality 63/100

Security audit verdict: SAFE · quality 63/100

No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →

by kyotofin · AI Tool · ★ 288

Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h

🔒 Is tax-doc-classifier safe to install? View the security audit →

About tax-doc-classifier

tax-doc-classifier A tax document classifier built with Jev. We ingest thousands of tax documents using an LLM pipeline built last tax season. Jev classifies 100% of our tax document corpus at $0.001 per page — 34× cheaper and 6× faster than that LLM setup. This is the classifier, open sourced. One request per page. The page's text goes to Jev, TypeSafe's decision model, which returns a probability over 261 IRS forms and 7 page kinds instead of text. No model is trained and nothing is hosted: the classifier is a JSON file describing each form, generated from the IRS's own PDFs. No model is trained and nothing is hosted. The classifier is a JSON file of form descriptions () and a decision model that reads a page and picks from them. The decision model is TypeSafe Jev, a System One model that returns a probability over options instead of text. Setup Node 20+ and pnpm. poppler — provides and , used to read PDF pages.

Quick Facts

Stars288
Forks29
LanguageTypeScript
CategoryAI Tool
LicenseApache-2.0
Quality Score63.0867165957086/100
Last Updated2026-09-20
Created2026-09-18
Platformsnode
Est. Tokens~5k

Compatible Skills

These tools work well together with tax-doc-classifier for enhanced workflows:

  • md.page — semantic(0.20)+complementary+same_lang+similar_pop+shared_platform (57%)

tax-doc-classifier alternative? Top 6 similar tools

Looking for a tax-doc-classifier alternative? If you're comparing tax-doc-classifier with other ai tool tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.

  • shaping-skills by rjs · ⭐ 1.4k

    Skills I use with Claude for shaping

  • pivotnacci by blackarrowsec · ⭐ 725

    A tool to make socks connections through HTTP agents

  • MedgeClaw by xjtulyc · ⭐ 668

    Open-source AI research assistant for biomedicine — chat to run RNA-seq, drug discovery, clinical analysis, an

  • SciToolAgent by HICAI-ZJU · ⭐ 416

    SciToolAgent: A Knowledge Graph-Driven Scientific Agent for Multi-Tool Integration

  • redesigned-pancake by Sfedfcv · ⭐ 256

    Skip to content github / docs Code Issues 80 Pull requests 35 Discussions Actions Projects 2 Security Insights

  • claude-code-voice-skill by abracadabra50 · ⭐ 169

    Skill to talk to Claude about your projects over the phone

More AI Tool Tools

Explore other popular ai tool tools:

View all AI Tool tools →

Popular TypeScript Agent Tools

Frequently Asked Questions

What is tax-doc-classifier?

tax-doc-classifier is Tax document page classifier built on Jev decisions. 100% strict accuracy across 261 IRS forms, ~$0.001 per page.. It is categorized as a AI Tool with 288 GitHub stars.

What programming language is tax-doc-classifier written in?

tax-doc-classifier is primarily written in TypeScript.

How do I install or use tax-doc-classifier?

You can find installation instructions and usage details in the tax-doc-classifier GitHub repository at github.com/kyotofin/tax-doc-classifier. The project has 288 stars and 29 forks, indicating an active community.

What license does tax-doc-classifier use?

tax-doc-classifier is released under the Apache-2.0 license, making it free to use and modify according to the license terms.

What are the best alternatives to tax-doc-classifier?

The top alternatives to tax-doc-classifier on Agent Skills Hub include shaping-skills, pivotnacci, MedgeClaw. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.

How this security grade is produced

Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.

The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.

Sources & who's responsible:

View on GitHub → Browse AI Tool tools