Eval-Driven Skill Design SKILL.md skillA

Eval-Driven Skill Design from s977043/river-review is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add s977043/river-review/Eval-Driven Skill Design@v0.2.0

Pins v0.2.0 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
s977043/river-review
Path
skills/upstream/eval-driven-skill-design/SKILL.md
Versions recorded
4
Latest version
v0.2.0, committed Oct 5, 2026
First version recorded
Jun 29, 2026
Size
10,625 bytes
Repository stars
7
License
MIT

新規・変更 skill の fixture/eval と、WITH/WITHOUT の paired ablation による限界寄与の評価可能性を確認し、activation と effectiveness を分離して案内する。.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

## Pattern declaration

Primary pattern: Reviewer
Secondary patterns: Tool Wrapper
Why: skill 追加・重要変更時に happy-path / guard fixture、eval 配線、paired ablation の評価可能性を診断し、存在・発火と実効性を混同しないための educator スキル。fixture-only / docs-only PR では起動しない。

## Goal / 目的

- 新規 `skills/**/SKILL.md` を追加する PR が **検出ケース (happy-path) と抑制ケース (guard) のペア** を fixture として持つことを確認する。
- `eval/promptfoo.yaml` 相当の評価設定または `tests/fixtures/repo-wide-eval/` への登録が用意されているかを確認する。
- 欠けている場合は `npm run eval:fixtures` / `npm run eval:repo-context` をどう通すかを案内する。
- 新規・重要変更された skill について、同じ case / runtime / model / effort で **WITH skill と WITHOUT skill(または旧版 baseline と candidate)** を比較できるか確認する。
- `installed` / `selected` / `fired` を effectiveness の証拠にせず、寄与が不明なら `INCONCLUSIVE` として追加 Evidence を要求する。

## Guidance

### Eval cycle 全体像 (#688 で整備済)

```text
new SKILL.md
  ├─ fixtures/
  │   ├─ <NN>-happy.md        # 検出されるべき diff(必須 1 件以上)
  │   └─ <NN>-guard.md        # 検出してはいけない diff(必須 1 件以上)
  ├─ golden/<NN>-*.md         # promptfoo similar 評価用の代表出力
  ├─ prompt/{system,user}.md  # promptfoo prompts:
  └─ eval/promptfoo.yaml      # contains / not-contains / llm-rubric / similar
```
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add s977043/river-review/Eval-Driven Skill Design@sha256:3b7b2f9197c0f37b

The line at the top of this page pins the label v0.2.0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_6zd3gzqqp4wazgue.svg)](https://markdownregistry.com/a/art_6zd3gzqqp4wazgue)

4 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 4 versions of this file, between Jun 29, 2026 and Oct 5, 2026, about one every 33 days.

versioncommittedcommitsizeaudit
v0.2.0 latest2026-10-05 4277a1a 10,625 BA view · diff
v0.1.02026-07-01 3696771 7,702 BA view · diff
v0.1.02026-06-29 45e66bc 7,719 BA view · diff
v0.1.02026-06-29 b988700 7,735 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (10625 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

s977043/river-review · 7 stars · license MIT · pushed 2026-11-08 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_6zd3gzqqp4wazgue
GET https://markdownregistry.com/api/v1/resolve?ref=s977043/river-review/Eval-Driven Skill Design
GET https://markdownregistry.com/api/v1/blob/3b7b2f9197c0f37bcb720bc361ba4e95ec214c15f92f59c01ca17b63d6a69c3d

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from s977043/river-review

ask-codex SKILL.md skill
s977043/river-review · .claude/skills/ask-codex/SKILL.md · Get a second opinion or independent review from Codex (OpenAI) via the local Codex CLI. Use when the user wants Codex…
git:20260724.d54eafd · audit A · 7 stars
river-review-discipline SKILL.md skill
s977043/river-review · .claude/skills/river-review-discipline/SKILL.md · AI 駆動開発のレビュー品質を安定化させるレビュー規律。レビュー対象を分類し、要件・設計・差分・検証・報告を分けて見て、根拠に基づき approved / needs_revision / blocked / rejected…
git:20260707.9c35a5d · audit A · 7 stars
skill-creator SKILL.md skill
s977043/river-review · .claude/skills/skill-creator/SKILL.md · Design a new Claude Code skill from a concrete use case and produce a repo-ready skill package. Use when the user asks…
git:20261004.95d1f7a · audit A · 7 stars
skill-ops-planner SKILL.md skill
s977043/river-review · .claude/skills/skill-ops-planner/SKILL.md · Create an operating policy and roadmap for a live portfolio of Claude Code skills. Use when the user asks to organize…
git:20260321.857d5ca · audit A · 7 stars
skill-optimizer SKILL.md skill
s977043/river-review · .claude/skills/skill-optimizer/SKILL.md · Evaluate and improve an existing Claude Code skill using explicit success criteria and small controlled changes. Use…
git:20261004.95d1f7a · audit A · 7 stars
AGENTS.md@.codex AGENTS.md file
s977043/river-review · .codex/AGENTS.md
git:20260923.5650e78 · audit A · 7 stars
AGENTS.md AGENTS.md file
s977043/river-review · AGENTS.md
git:20261008.d74129b · audit B · 7 stars
CLAUDE.md CLAUDE.md file
s977043/river-review · CLAUDE.md
git:20261016.787c99c · audit A · 7 stars
Sample Architecture Consistency Review SKILL.md skill
s977043/river-review · examples/skills/architecture-sample/SKILL.md · Checks design/ADR docs for consistency and missing decisions.
v0.1.0 · audit A · 7 stars
Sample Code Quality Pass SKILL.md skill
s977043/river-review · examples/skills/code-quality-sample/SKILL.md · Checks common code quality and maintainability risks.
v0.1.0 · audit A · 7 stars
Sample Test Coverage Review SKILL.md skill
s977043/river-review · examples/skills/test-review-sample/SKILL.md · Evaluates downstream tests for coverage and edge cases.
v0.1.0 · audit A · 7 stars
adversarial-review SKILL.md skill
s977043/river-review · plugin-dist/river-review/skills/agent-skills/adversarial-review/SKILL.md · 敵対的分析手法を統合したレビューの entry skill。認知バイアス対策の3手法 (Pre-mortem / War Game / Logic Torturing)と、宣言・主張と実態の乖離を突く claim-vs-actual…
v0.1.0 · audit A · 7 stars

Every file in s977043/river-review

Browse by kind, by grade A, or by owner.