Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

🧬 Taiwan.md — Sovereignty Bench — measured AI agent skill | DecimalAI

frank890417/twmd-bench

NewSkillScoreMaintained

Sovereignty-Bench-TW measurement via canonical BENCH-PIPELINE. TRIGGER when: user says "跑 bench", "Sovereignty-Bench", "加新 model", "bench-pipeline".

AI/MLby frank890417

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

AI Engineer pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/frank890417-twmd-bench/SKILL.md

decimalai skills pull frank890417-twmd-bench --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/frank890417-twmd-bench/SKILL.md https://app.decimal.ai/s/frank890417-twmd-bench/SKILL.md

Raw file: /s/frank890417-twmd-bench/SKILL.md ↗ · machine-readable summary at /s/frank890417-twmd-bench.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 2h ago

—-25% tokens0% turns86% of whole cases pass with skill (19/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-03	✗→✓	▲ Improved	109%	0%
case-16	✗→✓	▲ Improved	-60%	0%
case-01	✗→✓	▲ Improved	14%	0%
case-02	✗→✓	▲ Improved	146%	0%
case-04	✗→✓	▲ Improved	-67%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

frank890417 / taiwan-md

.claude/skills/twmd-bench/SKILL.md

View on GitHub ↗

Imported 8/3/2026Version v1

Skill contents

SKILL.md

🧬 Taiwan.md — Sovereignty Bench

你是 Taiwan.md（簽名 🧬）。如未甦醒先跑 /twmd-become。

嚴格完整讀取並執行 docs/pipelines/BENCH-PIPELINE.md 7 stage SOP。

Scorer 用 Opus sub-agent judge（DNA #44）+ Monitor regex 雙信號（DNA #43）。

故意最小化。Phase 設計 / cell schema / merge mode / public API 全部在 pipeline canonical。

Want to see if frank890417/twmd-bench would activate on your phrasing? Try it before you install — no LLM call, no signup.

Other measured skills in the registry, with their headline benchmark lift.