Soul Grader

Rubric-driven grader/rewriter for Hermes Agent SOUL.md identity files.

Tested · Works

Test report

Verdict
Tested · Works
Score
9.6/10
Tested
Jul 14, 2026
Environment
Claude Code 2.x (agent harness)
Upstream re-checked
Aug 10, 2026 · 82478e0

Fed it a toy SOUL.md full of generic filler ('be proactive', 'use best practices', 'never hallucinate') and it caught every phrase verbatim against its own banned-wording list, then scored it ~10/100 'Needs rewrite' with a full category breakdown instead of a vague thumbs-up.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

  • Installs cleanly 5/5
  • Triggers reliably 5/5
  • Output vs. baseline 9/10
  • Docs & honesty 5/5

What Soul Grader does

Grades, reviews, or rewrites a Hermes Agent SOUL.md against a bundled 11-category, 100-point rubric with automatic-fail checks (secrets, false claims, missing approval gates). Triggers on requests to grade, score, audit, approve, or rewrite a SOUL.md, or to decide whether an agent identity is deploy-ready.

How to install Soul Grader

git clone https://github.com/cobibean/soul-grader-skill
cd soul-grader-skill
mkdir -p ~/.claude/skills
cp -r . ~/.claude/skills/soul-grader

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Soul Grader

  • /soul-grader Rubric-driven grader/rewriter for Hermes Agent SOUL.md identity files.

It also activates on plain-language prompts like these:

  • Grade this SOUL.md against the 100-point rubric
  • Is this agent identity file actually deploy-ready?
  • Rewrite this SOUL.md to fix the automatic-fail issues

Frequently asked questions

Is the Soul Grader skill free?
Yes. The skill itself is free from cobibean/soul-grader-skill. SkillProof publishes the install command and an independent test verdict at no cost.
Does Soul Grader work with Claude Code?
We tested it with Claude Code 2.x (agent harness) on Jul 14, 2026. Verdict: Tested · Works. Fed it a toy SOUL.md full of generic filler ('be proactive', 'use best practices', 'never hallucinate') and it caught every phrase verbatim against its own banned-wording list, then scored it ~10/100 'Needs rewrite' with a full category breakdown instead of a vague thumbs-up.
What is the Soul Grader SkillProof Score?
9.6/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 9/10, docs & honesty 5/5.
How do I install Soul Grader?
Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Soul Grader with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.