What does the context bar show?

A visual bar showing the percentage of the 200K context window that is currently used.

What does the diagnosis output include?

It includes Weight (session size and message count), Tokens (exact or estimated), Context bar, Vital signs, Message type breakdown, Top 10 largest messages, and Estimated savings by prescription.

When should I apply a prescription and how are decisions made?

If context usage exceeds 60%, surface the result and recommend /cozempic:treat with a prescription. Use size thresholds: under 5MB for gentle, 5-20MB for standard, and over 20MB for aggressive.

diagnose

Scanned

npx machina-cli add skill Ruya-AI/cozempic/diagnose --openclaw

Files (1)

SKILL.md

1.1 KB

Run a diagnosis on the current session:

cozempic current --diagnose

The output includes:

Weight: total session size in bytes and message count
Tokens: exact token count (from usage data) or heuristic estimate
Context bar: visual bar showing % of the 200K context window used
Vital signs: progress ticks, file history snapshots, system reminders, thinking content, signatures, tool results
Message type breakdown: bytes per message type
Top 10 largest messages: biggest bloat contributors
Estimated savings by prescription: what gentle/standard/aggressive would save

Always surface the token count and context % to the user. If context is above 60%, suggest running /cozempic:treat with a prescription recommendation:

Under 5MB: gentle
5-20MB: standard
Over 20MB: aggressive

Source

git clone https://github.com/Ruya-AI/cozempic/blob/main/plugin/skills/diagnose/SKILL.mdView on GitHub

Overview

Diagnose analyzes the current Claude Code session to reveal token counts, context usage, and bloat contributors. It helps you understand session size, context window pressure, and where to trim for optimal performance.

How This Skill Works

Run the diagnosis with cozempic current --diagnose to generate a report that includes Weight, Tokens, Context bar, Vital signs, and a Message type breakdown. The tool always surfaces the token count and context % to the user; if context exceeds 60%, it suggests running /cozempic:treat with a prescription based on the session size (Under 5MB gentle, 5-20MB standard, Over 20MB aggressive).

When to Use It

User asks about session size or context usage
Context window feels full or near capacity (high context %)
You need to identify top bloat contributors (Top 10 largest messages)
You want estimated savings by prescription (gentle/standard/aggressive)
Explaining diagnostic results to a non-technical user

Quick Start

Step 1: Run cozempic current --diagnose
Step 2: Review the report, focusing on Weight, Tokens, Context bar, and Top 10 largest messages
Step 3: If context > 60%, run /cozempic:treat with a prescription (gentle, standard, or aggressive) based on size: Under 5MB gentle; 5-20MB standard; Over 20MB aggressive

Best Practices

Always surface token count and context % in every diagnosis report
Highlight the Top 10 largest messages to pinpoint bloat sources
Provide concrete next steps and a prescription recommendation when needed
Cross-check Weight with the actual message count and content types
Reference the 200K context window to interpret the context bar accurately

Example Use Cases

A user asks why their Claude Code session feels large or slow.
Context bar shows 68% usage; you recommend running /cozempic:treat with a standard prescription.
Diagnosis reveals long tool outputs and historical summaries as major contributors.
You compute and present estimated MB savings for gentle vs standard vs aggressive prescriptions.
You explain Weight, Tokens, and Vital signs to a teammate to justify actions.

Frequently Asked Questions

Add this skill to your agents