Skip to content

Claude Coach

Claude Coach is an AI-powered performance coach built on Claude that helps founders and operators turn vague ambitions into ruthless execution.

Code Video Cerebral Valley

Core trick

Persistent project memory feeds a persona prompt that refines goals, challenges excuses, and outputs dated action items.

Tags

productivity

Description

Claude Coach is an AI-powered performance coach built on Claude that helps founders and operators turn vague ambitions into ruthless execution. It combines goal-structuring, accountability check-ins, and action-plan generation with a harsh-but-funny “legendary hard-ass coach” personality to keep users engaged instead of ghosting their own plans. Under the hood, it uses structured prompts and project memory to iteratively refine user goals, challenge excuses, and output concrete next steps with deadlines, optimized for fast iteration in a startup environment. The result is a zero-friction “tough love” coach that feels human enough to be entertaining, but systematic enough to drive measurable progress over a single hackathon weekend and beyond.
anilkatti/claudecoach★ 00 forksPythonCreated Jun 13, 2026 · last push Jun 27, 2026

README excerpt

# ClaudeCoach

A Claude Code **plugin that coaches you on how you use Claude.** It learns who you are
and what your project is from your own session history, then helps you get more out of
Claude — what to install, what to trim, what to change — and applies the changes you
approve.

It's for **everyone** who works with Claude Code — engineers, analysts, accountants,
writers, lawyers. "Work" means any artifact: a spreadsheet, a document, a dataset, a
codebase. Nothing here assumes you write code.

Everything runs **locally and privately.** Your sessions are read on your machine,
secrets are scrubbed before any model sees them, and nothing is uploaded.

## How it works

ClaudeCoach is three skills you run in order — a **sensor** that collects, a **coach**
that judges, and an **executor** that acts. Each stays in its lane: the sensor never
recommends, the coach never senses and never touches your files, the executor only
applies what the coach proposed and you approved.

```
  your past sessions ─┐
  installed skills    ─┤  /profile-builder   ─▶  ~/.claude/profiles/<slug>/
  config health       ─┘   (SENSE)                project + user profile, profile.md
                                                          │
                                                          ▼
                          /recommend-actions   ─▶  actions.json + actions.html
                           (RECOMMEND)               prioritized, evidence-cited

Team