Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Prompt Sensei

A quiet prompt mentor that scores each prompt by stage and appends one coaching line to every answer in Claude Code or Codex.

Dev & CodingIntermediate★ 110⑂ 10AI score 10/10Last updated: Jul 13, 2026

What it does

  • Observe mode: scores the prompts you write during a session and appends exactly one Sensei line to each final answer, e.g. > **[[Sensei: 68/100 · Diagnosis; Tip: ...]]()**.
  • Stage-aware grading: classifies each prompt as Exploration, Diagnosis, Execution, Verification, Reusable workflow, or Action, then grades only the dimensions that matter for that stage — exploratory prompts are not punished for missing implementation detail.
  • Seven dimensions: Goal Clarity, Context Completeness, Input Boundaries, Constraints, Output Format, Verification, and Privacy/Safety, averaged into a 0–100 readiness score with grade labels.
  • Improve mode: shows what is missing and returns a minimally rewritten, copy-ready prompt plus exactly one habit to practice next.
  • Lookback & reports: with separate consent, analyzes selected local Claude Code/Codex history and prints a private Markdown summary. Data lives only in ~/.prompt-sensei/; raw prompt text is not stored.

Who it's for

  • Engineers who use AI coding agents daily and keep re-explaining themselves
  • Anyone who wants to fix prompting habits one at a time instead of reading theory
  • Team leads who want a shared standard for context, scope, and verification in requests
  • People who prefer unobtrusive feedback: one line, at the end, nothing else

Examples

  1. /prompt-sensei observe — start coaching; a short prompt like "login is broken" gets 48/100 · Diagnosis; Tip: add expected behavior and actual behavior before asking for a fix.
  2. /prompt-sensei improve "refactor this" — prints stage, score, missing pieces (file path, scope boundary, verification command) and a rewritten prompt you can paste back.
  3. /prompt-sensei report — shows your private summary of the habits you skip most often across sessions.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/chengzhongwei/Prompt-sensei/HEAD/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the SKILL.m folder from the GitHub repo chengzhongwei/Prompt-sensei into my ~/.claude/skills/prompt-sensei/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/chengzhongwei/Prompt-sensei.git ~/.claude/skills/prompt-sensei && cd ~/.claude/skills/prompt-sensei && npm install && npm run build

⚠ This is a third-party skill. Check the source repository before installing.

  1. Make sure Node.js is installed: run node -v (install the LTS build from nodejs.org if missing).
  2. Clone the skill into your skills folder: git clone https://github.com/chengzhongwei/Prompt-sensei.git ~/.claude/skills/prompt-sensei (Codex users should use ~/.codex/skills/prompt-sensei instead.)
  3. Build it: cd ~/.claude/skills/prompt-sensei && npm install && npm run build, then confirm dist/scripts/observe.js and report.js exist.
  4. Restart Claude Code and run /prompt-sensei setup to configure consent, auto-start scope, and optional redacted previews. On Codex, inspect hook command paths with /hooks before trusting them.
  5. Run /prompt-sensei observe to start coaching and /prompt-sensei stop to pause it.
  6. Use /prompt-sensei clear to delete local data and /prompt-sensei update to pull and rebuild the latest version.
View source on GitHub ↗License: Apache-2.0